All of lore.kernel.org
 help / color / mirror / Atom feed
From: Peter Maydell <peter.maydell@linaro.org>
To: Eric Auger <eric.auger@redhat.com>
Cc: eric.auger.pro@gmail.com, qemu-arm <qemu-arm@nongnu.org>,
	QEMU Developers <qemu-devel@nongnu.org>,
	Prem Mallappa <prem.mallappa@gmail.com>,
	Alex Williamson <alex.williamson@redhat.com>,
	Tomasz Nowicki <tn@semihalf.com>,
	"Michael S. Tsirkin" <mst@redhat.com>,
	Christoffer Dall <christoffer.dall@linaro.org>,
	Bharat Bhushan <bharat.bhushan@nxp.com>,
	Jean-Philippe Brucker <jean-philippe.brucker@arm.com>,
	"Edgar E. Iglesias" <edgar.iglesias@gmail.com>,
	linuc.decode@gmail.com, Peter Xu <peterx@redhat.com>
Subject: Re: [Qemu-devel] [PATCH v9 02/14] hw/arm/smmu-common: IOMMU memory region and address space setup
Date: Tue, 6 Mar 2018 14:08:07 +0000	[thread overview]
Message-ID: <CAFEAcA-uMZwR8HfVDx0Rau9G2unMypWJEL0o3ACPHxCqZ+g=_Q@mail.gmail.com> (raw)
In-Reply-To: <1518893216-9983-3-git-send-email-eric.auger@redhat.com>

On 17 February 2018 at 18:46, Eric Auger <eric.auger@redhat.com> wrote:
> We enumerate all the PCI devices attached to the SMMU and
> initialize an associated IOMMU memory region and address space.
> This happens on SMMU base instance init.
>
> Those info are stored in SMMUDevice objects. The devices are
> grouped according to the PCIBus they belong to. A hash table
> indexed by the PCIBus poinet is used. Also an array indexed by

"pointer".

> the bus number allows to find the list of SMMUDevices.
>
> Signed-off-by: Eric Auger <eric.auger@redhat.com>
>
> ---
> v8 -> v9:
> - fix key value for lookup
>
> v7 -> v8:
> - introduce SMMU_MAX_VA_BITS
> - use PCI bus handle as a key
> - do not clear s->smmu_as_by_bus_num
> - use g_new0 instead of g_malloc0
> - use primary_bus field
> ---
>  hw/arm/smmu-common.c         | 59 ++++++++++++++++++++++++++++++++++++++++++++
>  include/hw/arm/smmu-common.h |  6 +++++
>  2 files changed, 65 insertions(+)
>
> diff --git a/hw/arm/smmu-common.c b/hw/arm/smmu-common.c
> index 86a5aab..d0516dc 100644
> --- a/hw/arm/smmu-common.c
> +++ b/hw/arm/smmu-common.c
> @@ -28,12 +28,71 @@
>  #include "qemu/error-report.h"
>  #include "hw/arm/smmu-common.h"
>
> +SMMUPciBus *smmu_find_as_from_bus_num(SMMUState *s, uint8_t bus_num)
> +{
> +    SMMUPciBus *smmu_pci_bus = s->smmu_as_by_bus_num[bus_num];

Variable name suggests this is a table of AddressSpaces indexed by
bus number, but the code says we're getting SMMUPCIBus objects from it?

> +
> +    if (!smmu_pci_bus) {
> +        GHashTableIter iter;
> +
> +        g_hash_table_iter_init(&iter, s->smmu_as_by_busptr);
> +        while (g_hash_table_iter_next(&iter, NULL, (void **)&smmu_pci_bus)) {
> +            if (pci_bus_num(smmu_pci_bus->bus) == bus_num) {
> +                s->smmu_as_by_bus_num[bus_num] = smmu_pci_bus;
> +                return smmu_pci_bus;
> +            }
> +        }

Why do we populate this hashtable lazily rather than when we
put the SMMUPciBus in the smmu_as_by_busptr table? Do we
expect this function not to ordinarily be called?

> +    }
> +    return smmu_pci_bus;
> +}
> +
> +static AddressSpace *smmu_find_add_as(PCIBus *bus, void *opaque, int devfn)
> +{
> +    SMMUState *s = opaque;
> +    SMMUPciBus *sbus = g_hash_table_lookup(s->smmu_as_by_busptr, bus);
> +    SMMUDevice *sdev;
> +
> +    if (!sbus) {
> +        sbus = g_malloc0(sizeof(SMMUPciBus) +
> +                         sizeof(SMMUDevice *) * SMMU_PCI_DEVFN_MAX);
> +        sbus->bus = bus;
> +        g_hash_table_insert(s->smmu_as_by_busptr, bus, sbus);
> +    }
> +
> +    sdev = sbus->pbdev[devfn];
> +    if (!sdev) {
> +        char *name = g_strdup_printf("%s-%d-%d",
> +                                     s->mrtypename,
> +                                     pci_bus_num(bus), devfn);
> +        sdev = sbus->pbdev[devfn] = g_new0(SMMUDevice, 1);
> +
> +        sdev->smmu = s;
> +        sdev->bus = bus;
> +        sdev->devfn = devfn;
> +
> +        memory_region_init_iommu(&sdev->iommu, sizeof(sdev->iommu),
> +                                 s->mrtypename,
> +                                 OBJECT(s), name, 1ULL << SMMU_MAX_VA_BITS);
> +        address_space_init(&sdev->as,
> +                           MEMORY_REGION(&sdev->iommu), name);

This leaks the memory pointed to by name, doesn't it?
(address_space_init() takes a copy of the name string, so you want
to g_free(name) here.)

> +    }
> +
> +    return &sdev->as;
> +}
> +
>  static void smmu_base_realize(DeviceState *dev, Error **errp)
>  {
>      SMMUState *s = ARM_SMMU(dev);
>
>      s->configs = g_hash_table_new_full(NULL, NULL, NULL, g_free);
>      s->iotlb = g_hash_table_new_full(NULL, NULL, NULL, g_free);
> +    s->smmu_as_by_busptr = g_hash_table_new(NULL, NULL);
> +
> +    if (s->primary_bus) {
> +        pci_setup_iommu(s->primary_bus, smmu_find_add_as, s);
> +    } else {
> +        error_setg(errp, "SMMU is not attached to any PCI bus!");
> +    }
>  }
>
>  static void smmu_base_reset(DeviceState *dev)
> diff --git a/include/hw/arm/smmu-common.h b/include/hw/arm/smmu-common.h
> index 8a9d931..aee96c2 100644
> --- a/include/hw/arm/smmu-common.h
> +++ b/include/hw/arm/smmu-common.h
> @@ -121,4 +121,10 @@ typedef struct {
>  #define ARM_SMMU_GET_CLASS(obj)                              \
>      OBJECT_GET_CLASS(SMMUBaseClass, (obj), TYPE_ARM_SMMU)
>
> +SMMUPciBus *smmu_find_as_from_bus_num(SMMUState *s, uint8_t bus_num);
> +
> +static inline uint16_t smmu_get_sid(SMMUDevice *sdev)
> +{
> +    return  ((pci_bus_num(sdev->bus) & 0xff) << 8) | sdev->devfn;
> +}

Can we have at least brief documentation comments for any
new functions or prototypes added to header files, please?

>  #endif  /* HW_ARM_SMMU_COMMON */
> --

thanks
-- PMM

  reply	other threads:[~2018-03-06 14:08 UTC|newest]

Thread overview: 63+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2018-02-17 18:46 [Qemu-devel] [PATCH v9 00/14] ARM SMMUv3 Emulation Support Eric Auger
2018-02-17 18:46 ` [Qemu-devel] [PATCH v9 01/14] hw/arm/smmu-common: smmu base device and datatypes Eric Auger
2018-03-06 12:09   ` Peter Maydell
2018-03-06 15:01     ` Auger Eric
2018-02-17 18:46 ` [Qemu-devel] [PATCH v9 02/14] hw/arm/smmu-common: IOMMU memory region and address space setup Eric Auger
2018-03-06 14:08   ` Peter Maydell [this message]
2018-03-06 14:47     ` Auger Eric
2018-03-06 14:49       ` Peter Maydell
2018-02-17 18:46 ` [Qemu-devel] [PATCH v9 03/14] hw/arm/smmu-common: VMSAv8-64 page table walk Eric Auger
2018-03-06 19:43   ` Peter Maydell
2018-03-07 16:23     ` Auger Eric
2018-03-07 16:35       ` Peter Maydell
2018-03-08 18:56         ` Auger Eric
2018-03-08 19:01           ` Peter Maydell
2018-02-17 18:46 ` [Qemu-devel] [PATCH v9 04/14] hw/arm/smmuv3: Skeleton Eric Auger
2018-03-08 14:27   ` Peter Maydell
2018-03-09 13:19     ` Auger Eric
2018-03-09 13:37       ` Peter Maydell
2018-03-09 13:49         ` Auger Eric
2018-02-17 18:46 ` [Qemu-devel] [PATCH v9 05/14] hw/arm/smmuv3: Wired IRQ and GERROR helpers Eric Auger
2018-03-08 17:49   ` Peter Maydell
2018-03-09 14:03     ` Auger Eric
2018-03-09 14:18       ` Peter Maydell
2018-03-09 14:50         ` Auger Eric
2018-02-17 18:46 ` [Qemu-devel] [PATCH v9 06/14] hw/arm/smmuv3: Queue helpers Eric Auger
2018-03-08 18:28   ` Peter Maydell
2018-03-09 16:43     ` Auger Eric
2018-02-17 18:46 ` [Qemu-devel] [PATCH v9 07/14] hw/arm/smmuv3: Implement MMIO write operations Eric Auger
2018-03-08 18:37   ` Peter Maydell
2018-03-09 16:42     ` Auger Eric
2018-02-17 18:46 ` [Qemu-devel] [PATCH v9 08/14] hw/arm/smmuv3: Event queue recording helper Eric Auger
2018-03-08 18:39   ` Peter Maydell
2018-03-09 17:16     ` Auger Eric
2018-02-17 18:46 ` [Qemu-devel] [PATCH v9 09/14] hw/arm/smmuv3: Implement translate callback Eric Auger
2018-03-09 18:46   ` Peter Maydell
2018-03-12 10:38     ` Eric Auger
2018-02-17 18:46 ` [Qemu-devel] [PATCH v9 10/14] hw/arm/smmuv3: Abort on vfio or vhost case Eric Auger
2018-03-08 19:06   ` Peter Maydell
2018-03-09 17:53     ` Auger Eric
2018-03-09 17:59       ` Peter Maydell
2018-03-12 10:53         ` Eric Auger
2018-03-12 11:10           ` Peter Maydell
2018-03-12 15:01             ` Auger Eric
2018-02-17 18:46 ` [Qemu-devel] [PATCH v9 11/14] target/arm/kvm: Translate the MSI doorbell in kvm_arch_fixup_msi_route Eric Auger
2018-03-12 11:59   ` Peter Maydell
2018-03-12 15:16     ` Auger Eric
2018-03-13 13:37       ` Paolo Bonzini
2018-03-15  9:45         ` Auger Eric
2018-02-17 18:46 ` [Qemu-devel] [PATCH v9 12/14] hw/arm/virt: Add SMMUv3 to the virt board Eric Auger
2018-03-12 12:46   ` Peter Maydell
2018-03-12 15:01     ` Auger Eric
2018-03-12 15:05       ` Peter Maydell
2018-02-17 18:46 ` [Qemu-devel] [PATCH v9 13/14] hw/arm/virt-acpi-build: Add smmuv3 node in IORT table Eric Auger
2018-03-12 12:48   ` Peter Maydell
2018-03-19 14:32     ` Shannon Zhao
2018-03-19 20:50       ` Auger Eric
2018-02-17 18:46 ` [Qemu-devel] [PATCH v9 14/14] hw/arm/virt: Handle iommu in 2.12 machine type Eric Auger
2018-03-12 12:56   ` Peter Maydell
2018-03-12 15:01     ` Auger Eric
2018-02-27 19:02 ` [Qemu-devel] [PATCH v9 00/14] ARM SMMUv3 Emulation Support Peter Maydell
2018-02-28  8:44   ` Auger Eric
2018-03-12 12:58     ` Peter Maydell
2018-03-12 15:22       ` Auger Eric

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='CAFEAcA-uMZwR8HfVDx0Rau9G2unMypWJEL0o3ACPHxCqZ+g=_Q@mail.gmail.com' \
    --to=peter.maydell@linaro.org \
    --cc=alex.williamson@redhat.com \
    --cc=bharat.bhushan@nxp.com \
    --cc=christoffer.dall@linaro.org \
    --cc=edgar.iglesias@gmail.com \
    --cc=eric.auger.pro@gmail.com \
    --cc=eric.auger@redhat.com \
    --cc=jean-philippe.brucker@arm.com \
    --cc=linuc.decode@gmail.com \
    --cc=mst@redhat.com \
    --cc=peterx@redhat.com \
    --cc=prem.mallappa@gmail.com \
    --cc=qemu-arm@nongnu.org \
    --cc=qemu-devel@nongnu.org \
    --cc=tn@semihalf.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 an external index of several public inboxes,
see mirroring instructions on how to clone and mirror
all data and code used by this external index.