[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]
Re: [Qemu-devel] [RFC v1 3/3] intel_iommu: add scalable-mode option to m
From: |
Peter Xu |
Subject: |
Re: [Qemu-devel] [RFC v1 3/3] intel_iommu: add scalable-mode option to make scalable mode work |
Date: |
Tue, 12 Feb 2019 14:46:29 +0800 |
User-agent: |
Mutt/1.10.1 (2018-07-13) |
On Wed, Jan 30, 2019 at 01:09:13PM +0800, Yi Sun wrote:
> This patch adds an option to provide flexibility for user to expose
> Scalable Mode to guest. User could expose Scalable Mode to guest by
> the config as below:
>
> "-device intel-iommu,caching-mode=on,scalable-mode=on"
>
> The Linux iommu driver has supported scalable mode. Please refer below
> patch set:
>
> https://www.spinics.net/lists/kernel/msg2985279.html
>
> Signed-off-by: Liu, Yi L <address@hidden>
> Signed-off-by: Yi Sun <address@hidden>
> ---
> hw/i386/intel_iommu.c | 22 ++++++++++++++++++++++
> hw/i386/intel_iommu_internal.h | 6 ++++++
> include/hw/i386/intel_iommu.h | 3 ++-
> 3 files changed, 30 insertions(+), 1 deletion(-)
>
> diff --git a/hw/i386/intel_iommu.c b/hw/i386/intel_iommu.c
> index 3664a00..447fdf3 100644
> --- a/hw/i386/intel_iommu.c
> +++ b/hw/i386/intel_iommu.c
> @@ -2492,6 +2492,17 @@ static bool vtd_process_inv_desc(IntelIOMMUState *s)
> }
> break;
>
> + /*
> + * TODO: the entity of below two cases will be implemented in future
> series.
> + * To make guest (which integrates scalable mode support patch set in
> + * iommu driver) work, just return true is enough so far.
When you repost, could you mention about what tests have you done? I
can think of some general usages:
Device matrix:
- virtio-net, vhost=on|off
- device assignment
Configuration matrix:
- legacy/scalable
- passthrough=on|off
I believe smoke test (like netperf a few seconds, or even ping) would
be enough, cover some (or all, which would be very nice to have :) of
above scenarios.
> + */
> + case VTD_INV_DESC_PC:
> + break;
> +
> + case VTD_INV_DESC_PIOTLB:
> + break;
> +
> case VTD_INV_DESC_WAIT:
> trace_vtd_inv_desc("wait", inv_desc.val[1], inv_desc.val[0]);
> if (!vtd_process_wait_desc(s, &inv_desc)) {
> @@ -3051,6 +3062,7 @@ static Property vtd_properties[] = {
> DEFINE_PROP_UINT8("aw-bits", IntelIOMMUState, aw_bits,
> VTD_HOST_ADDRESS_WIDTH),
> DEFINE_PROP_BOOL("caching-mode", IntelIOMMUState, caching_mode, FALSE),
> + DEFINE_PROP_BOOL("scalable-mode", IntelIOMMUState, scalable_mode, FALSE),
Let's start with x-scalable-mode? Less burden for all.
> DEFINE_PROP_BOOL("dma-drain", IntelIOMMUState, dma_drain, true),
> DEFINE_PROP_END_OF_LIST(),
> };
> @@ -3583,6 +3595,16 @@ static void vtd_init(IntelIOMMUState *s)
> s->cap |= VTD_CAP_CM;
> }
>
> + /* TODO: read cap/ecap from host to decide which cap to be exposed. */
> + if (s->scalable_mode) {
> + if (!s->caching_mode) {
> + error_report("Need to set caching-mode for scalable mode");
Could I ask why?
> + exit(1);
> + }
> + s->cap |= VTD_CAP_DWD | VTD_CAP_DRD;
Draining is supported now so this line is not needed.
> + s->ecap |= VTD_ECAP_SMTS | VTD_ECAP_SRS | VTD_ECAP_SLTS;
> + }
> +
> vtd_reset_caches(s);
>
> /* Define registers with default values and bit semantics */
> diff --git a/hw/i386/intel_iommu_internal.h b/hw/i386/intel_iommu_internal.h
> index 2a753c5..b01953a 100644
> --- a/hw/i386/intel_iommu_internal.h
> +++ b/hw/i386/intel_iommu_internal.h
> @@ -190,7 +190,9 @@
> #define VTD_ECAP_EIM (1ULL << 4)
> #define VTD_ECAP_PT (1ULL << 6)
> #define VTD_ECAP_MHMV (15ULL << 20)
> +#define VTD_ECAP_SRS (1ULL << 31)
> #define VTD_ECAP_SMTS (1ULL << 43)
> +#define VTD_ECAP_SLTS (1ULL << 46)
>
> /* CAP_REG */
> /* (offset >> 4) << 24 */
> @@ -209,6 +211,8 @@
> #define VTD_CAP_DRAIN_READ (1ULL << 55)
> #define VTD_CAP_DRAIN (VTD_CAP_DRAIN_READ |
> VTD_CAP_DRAIN_WRITE)
> #define VTD_CAP_CM (1ULL << 7)
> +#define VTD_CAP_DWD (1ULL << 54)
> +#define VTD_CAP_DRD (1ULL << 55)
These can be dropped too (see VTD_CAP_DRAIN above).
>
> /* Supported Adjusted Guest Address Widths */
> #define VTD_CAP_SAGAW_SHIFT 8
> @@ -340,6 +344,8 @@ typedef union VTDInvDesc VTDInvDesc;
> #define VTD_INV_DESC_IEC 0x4 /* Interrupt Entry Cache
> Invalidate Descriptor */
> #define VTD_INV_DESC_WAIT 0x5 /* Invalidation Wait Descriptor
> */
> +#define VTD_INV_DESC_PIOTLB 0x6 /* PASID-IOTLB Invalidate Desc */
> +#define VTD_INV_DESC_PC 0x7 /* PASID-cache Invalidate Desc */
> #define VTD_INV_DESC_NONE 0 /* Not an Invalidate Descriptor
> */
>
> /* Masks for Invalidation Wait Descriptor*/
> diff --git a/include/hw/i386/intel_iommu.h b/include/hw/i386/intel_iommu.h
> index a5da139..a04fad6 100644
> --- a/include/hw/i386/intel_iommu.h
> +++ b/include/hw/i386/intel_iommu.h
> @@ -219,7 +219,8 @@ struct IntelIOMMUState {
> uint8_t womask[DMAR_REG_SIZE]; /* WO (write only - read returns 0) */
> uint32_t version;
>
> - bool caching_mode; /* RO - is cap CM enabled? */
> + bool caching_mode; /* RO - is cap CM enabled? */
> + bool scalable_mode; /* RO - is Scalable Mode supported? */
>
> dma_addr_t root; /* Current root table pointer */
> bool root_extended; /* Type of root table (extended or not)
> */
> --
> 1.9.1
>
Regards,
--
Peter Xu
- Re: [Qemu-devel] [RFC v1 3/3] intel_iommu: add scalable-mode option to make scalable mode work,
Peter Xu <=