qemu-devel.nongnu.org archive mirror
 help / color / mirror / Atom feed
From: Peter Xu <peterx@redhat.com>
To: marcel@redhat.com
Cc: qemu-devel@nongnu.org, ehabkost@redhat.com, mst@redhat.com,
	jasowang@redhat.com, pbonzini@redhat.com, imammedo@redhat.com,
	rth@twiddle.net, David Kiarie <davidkiarie4@gmail.com>
Subject: Re: [Qemu-devel] [PATCH 01/13] q35: add "int-remap" flag to enable intr
Date: Fri, 8 Apr 2016 15:30:16 +0800	[thread overview]
Message-ID: <20160408073016.GA26923@pxdev.xzpeter.org> (raw)
In-Reply-To: <56C993AE.7010703@gmail.com>

It's a long time from previous post... However I will start to pick
this up. Several questions on the comments...

On Sun, Feb 21, 2016 at 12:38:38PM +0200, Marcel Apfelbaum wrote:
> On 02/19/2016 05:30 AM, Peter Xu wrote:
> >One flag is added to specify whether to enable INTR for emulated
> >IOMMU. By default, interrupt remapping is not supportted. To enable it,
> >we should specify something like:
> >
> >$ qemu-system-x86_64 -M q35,iommu=on,int_remap=on
> 
> Hi Peter,
> 
> Please be aware that there is an AMD IOMMU emulation series on the list,
> so now we will have iommu=intel/amd/off.

I had a look on the AMD series. I see that x-iommu-type is
introduced to specify AMD IOMMUs. So maybe I should keep this
interface unchanged for now?

> 
> As far as I know we prefer int-remap instead of int_remap and also
> the "int" prefix reminds me integer, I would use ir=on or the clean interrupt-remapping=on.

I suppose "ir" is a little bit hard for people to think about
interrupt remaping, and "interrupt-remapping" might be too long. If
you would not mind, I'd like to use "intr" in next version.

[...]

> >+static bool machine_get_int_remap(Object *obj, Error **errp)
> >+{
> >+    MachineState *ms = MACHINE(obj);
> >+
> >+    return ms->int_remap;
> >+}
> 
> I am starting to wonder if "iommu" and now "interrupt-remapping" should be part
> of machine and not the pc machine. Both implementations we have are only for the PC machines.

Do you mean that "both implementation we have are for the machines"
rather than PC machines? It seems that both of us are modifying
machine_initfn() rather than pc_machine_initfn()? Am I missing
anything?

> 
> >+
> >+static void machine_set_int_remap(Object *obj, bool value, Error **errp)
> >+{
> >+    MachineState *ms = MACHINE(obj);
> >+
> >+    ms->int_remap = value;
> >+}
> >+
> >  static void machine_set_suppress_vmdesc(Object *obj, bool value, Error **errp)
> >  {
> >      MachineState *ms = MACHINE(obj);
> >@@ -461,6 +475,12 @@ static void machine_initfn(Object *obj)
> >      object_property_set_description(obj, "iommu",
> >                                      "Set on/off to enable/disable Intel IOMMU (VT-d)",
> >                                      NULL);
> >+    object_property_add_bool(obj, "int-remap",
> >+                             machine_get_int_remap,
> >+                             machine_set_int_remap, NULL);
> >+    object_property_set_description(obj, "int-remap",
> >+                                    "Set on/off to enable/disable Intel IOMMU INTR",
> >+                                    NULL);
> 
> 
> Here the same, I would rename int-remap to interrupt-remapping and keep in mind
> that would not be for Intel IOMMU only.

Will keep that in mind. Thanks.

However, we do not need to specify the type here, right? Since we
can specify the type of IOMMU via x-iommu-type, and the type of IR
will always follow the type of IOMMU.

> 
> >      object_property_add_bool(obj, "suppress-vmdesc",
> >                               machine_get_suppress_vmdesc,
> >                               machine_set_suppress_vmdesc, NULL);
> >diff --git a/hw/pci-host/q35.c b/hw/pci-host/q35.c
> >index 115fb8c..7cd4d87 100644
> >--- a/hw/pci-host/q35.c
> >+++ b/hw/pci-host/q35.c
> >@@ -434,13 +434,14 @@ static AddressSpace *q35_host_dma_iommu(PCIBus *bus, void *opaque, int devfn)
> >      return &vtd_as->as;
> >  }
> >
> >-static void mch_init_dmar(MCHPCIState *mch)
> >+static void mch_init_dmar(MCHPCIState *mch, bool intr_supported)
> >  {
> >      PCIBus *pci_bus = PCI_BUS(qdev_get_parent_bus(DEVICE(mch)));
> >
> >      mch->iommu = INTEL_IOMMU_DEVICE(qdev_create(NULL, TYPE_INTEL_IOMMU_DEVICE));
> >      object_property_add_child(OBJECT(mch), "intel-iommu",
> >                                OBJECT(mch->iommu), NULL);
> >+    mch->iommu->intr_supported = intr_supported;
> >      qdev_init_nofail(DEVICE(mch->iommu));
> >      sysbus_mmio_map(SYS_BUS_DEVICE(mch->iommu), 0, Q35_HOST_BRIDGE_IOMMU_ADDR);
> >
> >@@ -507,7 +508,8 @@ static void mch_realize(PCIDevice *d, Error **errp)
> >      }
> >      /* Intel IOMMU (VT-d) */
> >      if (object_property_get_bool(qdev_get_machine(), "iommu", NULL)) {
> >-        mch_init_dmar(mch);
> >+        mch_init_dmar(mch, object_property_get_bool(qdev_get_machine(),
> >+                                                    "int-remap", false));
> 
> 
> Here you can directly call qdev_get_machine()->interrupt-remapping instead of object_property_get_bool.

Will do.

Thanks!

-- peterx

  parent reply	other threads:[~2016-04-08  7:30 UTC|newest]

Thread overview: 43+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2016-02-19  3:30 [Qemu-devel] [PATCH 00/13] IOMMU: Enable interrupt remapping for Intel IOMMU Peter Xu
2016-02-19  3:30 ` [Qemu-devel] [PATCH 01/13] q35: add "int-remap" flag to enable intr Peter Xu
2016-02-21 10:38   ` Marcel Apfelbaum
2016-02-23  3:48     ` Peter Xu
2016-02-25 15:47       ` Marcel Apfelbaum
2016-04-08  7:30     ` Peter Xu [this message]
2016-04-11 10:07       ` Marcel Apfelbaum
2016-02-19  3:30 ` [Qemu-devel] [PATCH 02/13] acpi: enable INTR for DMAR report structure Peter Xu
2016-02-21 11:05   ` Marcel Apfelbaum
2016-04-08  8:07     ` Peter Xu
2016-02-19  3:30 ` [Qemu-devel] [PATCH 03/13] intel_iommu: allow queued invalidation for IR Peter Xu
2016-02-19  3:30 ` [Qemu-devel] [PATCH 04/13] intel_iommu: set IR bit for ECAP register Peter Xu
2016-02-19  3:30 ` [Qemu-devel] [PATCH 05/13] acpi: add DMAR scope definition for root IOAPIC Peter Xu
2016-02-21 11:38   ` Marcel Apfelbaum
2016-02-21 12:08     ` Marcel Apfelbaum
2016-02-21 13:40       ` Jan Kiszka
2016-02-21 15:54         ` Marcel Apfelbaum
2016-02-21 16:01           ` Jan Kiszka
2016-04-08  9:53             ` Peter Xu
2016-02-19  3:30 ` [Qemu-devel] [PATCH 06/13] intel_iommu: define interrupt remap table addr register Peter Xu
2016-02-19  3:30 ` [Qemu-devel] [PATCH 07/13] intel_iommu: handle interrupt remap enable Peter Xu
2016-02-19  3:30 ` [Qemu-devel] [PATCH 08/13] intel_iommu: define several structs for IOMMU IR Peter Xu
2016-02-19  3:30 ` [Qemu-devel] [PATCH 09/13] intel_iommu: provide helper function vtd_get_iommu Peter Xu
2016-02-19  3:30 ` [Qemu-devel] [PATCH 10/13] ioapic-common: add iommu for IOAPICCommonState Peter Xu
2016-02-19  3:30 ` [Qemu-devel] [PATCH 11/13] intel_iommu: add IR translation faults defines Peter Xu
2016-02-21 15:56   ` Marcel Apfelbaum
2016-04-08 10:03     ` Peter Xu
2016-02-19  3:30 ` [Qemu-devel] [PATCH 12/13] intel_iommu: ioapic: IR support for emulated IOAPIC Peter Xu
2016-02-19  3:30 ` [Qemu-devel] [PATCH 13/13] intel_iommu: Add support for PCI MSI remap Peter Xu
2016-02-19  6:46 ` [Qemu-devel] [PATCH 00/13] IOMMU: Enable interrupt remapping for Intel IOMMU Jan Kiszka
2016-02-19  7:43   ` Peter Xu
2016-02-19  8:34     ` Jan Kiszka
2016-02-19  9:29       ` Peter Xu
2016-02-19  9:58         ` Paolo Bonzini
2016-02-19 10:15           ` Jan Kiszka
2016-02-19 11:39             ` Peter Xu
2016-02-19 11:43               ` Jan Kiszka
2016-02-19 11:34           ` Peter Xu
2016-02-19 11:43             ` Jan Kiszka
2016-02-19 16:22               ` Radim Krčmář
2016-02-20 10:05         ` Jan Kiszka
2016-02-19 16:38   ` Radim Krčmář
2016-02-23  5:03     ` Peter Xu

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=20160408073016.GA26923@pxdev.xzpeter.org \
    --to=peterx@redhat.com \
    --cc=davidkiarie4@gmail.com \
    --cc=ehabkost@redhat.com \
    --cc=imammedo@redhat.com \
    --cc=jasowang@redhat.com \
    --cc=marcel@redhat.com \
    --cc=mst@redhat.com \
    --cc=pbonzini@redhat.com \
    --cc=qemu-devel@nongnu.org \
    --cc=rth@twiddle.net \
    /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).