On 07-Jul-21 7:25 AM, Ding, Xuan wrote:
Hi,
-----Original Message-----
From: Maxime Coquelin <maxime.coque...@redhat.com>
Sent: Tuesday, July 6, 2021 5:32 PM
To: Burakov, Anatoly <anatoly.bura...@intel.com>; Ding, Xuan
<xuan.d...@intel.com>; Xia, Chenbo <chenbo....@intel.com>; Thomas
Monjalon <tho...@monjalon.net>; David Marchand
<david.march...@redhat.com>
Cc: dev@dpdk.org; Hu, Jiayu <jiayu...@intel.com>; Pai G, Sunil
<sunil.pa...@intel.com>; Richardson, Bruce <bruce.richard...@intel.com>; Van
Haaren, Harry <harry.van.haa...@intel.com>; Liu, Yong <yong....@intel.com>;
Ma, WenwuX <wenwux...@intel.com>
Subject: Re: [dpdk-dev] [PATCH v5 1/2] vhost: enable IOMMU for async vhost
On 7/6/21 11:16 AM, Burakov, Anatoly wrote:
On 06-Jul-21 9:31 AM, Ding, Xuan wrote:
Hi Maxime,
-----Original Message-----
From: Maxime Coquelin <maxime.coque...@redhat.com>
Sent: Monday, July 5, 2021 8:46 PM
To: Burakov, Anatoly <anatoly.bura...@intel.com>; Ding, Xuan
<xuan.d...@intel.com>; Xia, Chenbo <chenbo....@intel.com>; Thomas
Monjalon <tho...@monjalon.net>; David Marchand
<david.march...@redhat.com>
Cc: dev@dpdk.org; Hu, Jiayu <jiayu...@intel.com>; Pai G, Sunil
<sunil.pa...@intel.com>; Richardson, Bruce <bruce.richard...@intel.com>;
Van Haaren, Harry <harry.van.haa...@intel.com>; Liu, Yong
<yong....@intel.com>; Ma, WenwuX <wenwux...@intel.com>
Subject: Re: [dpdk-dev] [PATCH v5 1/2] vhost: enable IOMMU for async
vhost
On 7/5/21 2:16 PM, Burakov, Anatoly wrote:
On 05-Jul-21 9:40 AM, Xuan Ding wrote:
The use of IOMMU has many advantages, such as isolation and address
translation. This patch extends the capbility of DMA engine to use
IOMMU if the DMA device is bound to vfio.
When set memory table, the guest memory will be mapped
into the default container of DPDK.
Signed-off-by: Xuan Ding <xuan.d...@intel.com>
---
doc/guides/prog_guide/vhost_lib.rst | 9 ++++++
lib/vhost/rte_vhost.h | 1 +
lib/vhost/socket.c | 9 ++++++
lib/vhost/vhost.h | 1 +
lib/vhost/vhost_user.c | 46
++++++++++++++++++++++++++++-
5 files changed, 65 insertions(+), 1 deletion(-)
diff --git a/doc/guides/prog_guide/vhost_lib.rst
b/doc/guides/prog_guide/vhost_lib.rst
index 05c42c9b11..c3beda23d9 100644
--- a/doc/guides/prog_guide/vhost_lib.rst
+++ b/doc/guides/prog_guide/vhost_lib.rst
@@ -118,6 +118,15 @@ The following is an overview of some key Vhost
API functions:
It is disabled by default.
+ - ``RTE_VHOST_USER_ASYNC_USE_VFIO``
+
+ In asynchronous data path, vhost liarary is not aware of which
driver
+ (igb_uio/vfio) the DMA device is bound to. Application should
pass
+ this flag to tell vhost library whether IOMMU should be
programmed
+ for guest memory.
+
+ It is disabled by default.
+
- ``RTE_VHOST_USER_NET_COMPLIANT_OL_FLAGS``
Since v16.04, the vhost library forwards checksum and gso
requests for
diff --git a/lib/vhost/rte_vhost.h b/lib/vhost/rte_vhost.h
index 8d875e9322..a766ea7b6b 100644
--- a/lib/vhost/rte_vhost.h
+++ b/lib/vhost/rte_vhost.h
@@ -37,6 +37,7 @@ extern "C" {
#define RTE_VHOST_USER_LINEARBUF_SUPPORT (1ULL << 6)
#define RTE_VHOST_USER_ASYNC_COPY (1ULL << 7)
#define RTE_VHOST_USER_NET_COMPLIANT_OL_FLAGS (1ULL << 8)
+#define RTE_VHOST_USER_ASYNC_USE_VFIO (1ULL << 9)
/* Features. */
#ifndef VIRTIO_NET_F_GUEST_ANNOUNCE
diff --git a/lib/vhost/socket.c b/lib/vhost/socket.c
index 5d0d728d52..77c722c86b 100644
--- a/lib/vhost/socket.c
+++ b/lib/vhost/socket.c
@@ -42,6 +42,7 @@ struct vhost_user_socket {
bool extbuf;
bool linearbuf;
bool async_copy;
+ bool async_use_vfio;
bool net_compliant_ol_flags;
/*
@@ -243,6 +244,13 @@ vhost_user_add_connection(int fd, struct
vhost_user_socket *vsocket)
dev->async_copy = 1;
}
+ if (vsocket->async_use_vfio) {
+ dev = get_device(vid);
+
+ if (dev)
+ dev->async_use_vfio = 1;
+ }
+
VHOST_LOG_CONFIG(INFO, "new device, handle is %d\n", vid);
if (vsocket->notify_ops->new_connection) {
@@ -879,6 +887,7 @@ rte_vhost_driver_register(const char *path,
uint64_t flags)
vsocket->extbuf = flags & RTE_VHOST_USER_EXTBUF_SUPPORT;
vsocket->linearbuf = flags &
RTE_VHOST_USER_LINEARBUF_SUPPORT;
vsocket->async_copy = flags & RTE_VHOST_USER_ASYNC_COPY;
+ vsocket->async_use_vfio = flags &
RTE_VHOST_USER_ASYNC_USE_VFIO;
vsocket->net_compliant_ol_flags = flags &
RTE_VHOST_USER_NET_COMPLIANT_OL_FLAGS;
if (vsocket->async_copy &&
diff --git a/lib/vhost/vhost.h b/lib/vhost/vhost.h
index 8078ddff79..fb775ce4ed 100644
--- a/lib/vhost/vhost.h
+++ b/lib/vhost/vhost.h
@@ -370,6 +370,7 @@ struct virtio_net {
int16_t broadcast_rarp;
uint32_t nr_vring;
int async_copy;
+ int async_use_vfio;
int extbuf;
int linearbuf;
struct vhost_virtqueue *virtqueue[VHOST_MAX_QUEUE_PAIRS *
2];
diff --git a/lib/vhost/vhost_user.c b/lib/vhost/vhost_user.c
index 8f0eba6412..f3703f2e72 100644
--- a/lib/vhost/vhost_user.c
+++ b/lib/vhost/vhost_user.c
@@ -45,6 +45,7 @@
#include <rte_common.h>
#include <rte_malloc.h>
#include <rte_log.h>
+#include <rte_vfio.h>
#include "iotlb.h"
#include "vhost.h"
@@ -141,6 +142,36 @@ get_blk_size(int fd)
return ret == -1 ? (uint64_t)-1 : (uint64_t)stat.st_blksize;
}
+static int
+async_dma_map(struct rte_vhost_mem_region *region, bool do_map)
+{
+ int ret = 0;
+ uint64_t host_iova;
+ host_iova = rte_mem_virt2iova((void
*)(uintptr_t)region->host_user_addr);
+ if (do_map) {
+ /* Add mapped region into the default container of DPDK. */
+ ret =
rte_vfio_container_dma_map(RTE_VFIO_DEFAULT_CONTAINER_FD,
+ region->host_user_addr,
+ host_iova,
+ region->size);
+ if (ret) {
+ VHOST_LOG_CONFIG(ERR, "DMA engine map failed\n");
+ return ret;
+ }
+ } else {
+ /* Remove mapped region from the default container of
DPDK. */
+ ret =
rte_vfio_container_dma_unmap(RTE_VFIO_DEFAULT_CONTAINER_FD,
+ region->host_user_addr,
+ host_iova,
+ region->size);
+ if (ret) {
+ VHOST_LOG_CONFIG(ERR, "DMA engine unmap failed\n");
+ return ret;
+ }
+ }
+ return ret;
+}
We've been discussing this off list with Xuan, and unfortunately
this is
a blocker for now.
Currently, the x86 IOMMU does not support partial unmap - the segments
have to be unmapped exactly the same addr/len as they were mapped. We
also concatenate adjacent mappings to prevent filling up the DMA
mapping
entry table with superfluous entries.
This means that, when two unrelated mappings are contiguous in memory
(e.g. if you map regions 1 and 2 independently, but they happen to be
sitting right next to each other in virtual memory), we cannot later
unmap one of them because, even though these are two separate
mappings
as far as kernel VFIO infrastructure is concerned, the mapping gets
compacted and looks like one single mapping to VFIO, so DPDK API will
not let us unmap region 1 without also unmapping region 2.
The proper fix for this problem would be to always map memory
page-by-page regardless of where it comes from (we already do that for
internal memory, but not for external). However, the reason this works
for internal memory is because when mapping internal memory segments,
*we know the page size*. For external memory segments, there is no such
guarantee, so we cannot deduce page size for a given memory segment,
and
thus can't map things page-by-page.
So, the proper fix for it would be to add page size to the VFIO DMA
API.
Unfortunately, it probably has to wait until 21.11 because it is an API
change.
The slightly hacky fix for this would be to forego user mem map
concatenation and trust that user is not going to do anything stupid,
and will not spam the VFIO DMA API without reason. I would rather
not go
down this road, but this could be an option in this case.
Thoughts?
Thanks Anatoly for the detailed description of the issue.
It may be possible to either create a versioned symbol for this API
change, or maybe even to have a temporary internal API.
But I think this series in its current form is not acceptable, so
waiting for v21.11 would be the best option (we may want to send the
deprecation notice in this release though).
In this series, I don't like the user application has to pass a flag to
state whether the DMA engine uses VFIO or not. AFAICT, this new revision
does not implement what was discussed in the previous one, i.e.
supporting both IOVA_AS_VA and IOVA_AS_PA.
Thanks for your comments. Here I hope to explain some questions:
1. Whether both IOVA_AS_VA and IOVA_AS_PA are supported now?
A: Both IOVA_AS_PA and IOVA_AS_VA are supported now. In this version, the
virtual address is replaced with iova address of mapped region, and
the iova
address is selected to program the IOMMU instead of virtual address only.
Good!
2. Why a flag is chosen to be passed by application?
A: Yes, as we discussed before, the rte_eal_iova_mode() API can be
used to
get the IOVA mode, so as to determine whether IOMMU should be
programmed.
However, in the implementation process, I found a problem. That is how to
distinguish the VFIO PA and IGB_UIO PA. Because for VFIO cases, we should
always program the IOMMU. While in IGB_UIO cases, it depends on IOMMU
capability of platform.
How does one program IOMMU with igb_uio? I was under impression that
igb_uio (and uio_pci_generic for that matter) does not provide such
facilities.
+1
Maybe some misunderstanding in this sentence here.
In our design, if rte_eal_vfio_is_enabled("vfio") is true, iommu will be
programmed.
True means vfio module is modprobed.
But there is an exception here, that is, even if vfio module is modprobed,
DPDK user still bind all the devices to igb_uio.
This situation can be distinguished in DPDK eal initialization, because the
resource mapping
is according to the driver loaded by each device(rte_pci_map_device).
But in our scenario, this judgment is somewhat weak. Because we cannot get
the device driver info in vhost library. I also think it is unreasonable for
vhost to
do this. Only trust that users will not use it like this. Thoughts for this
scenario?
I don't see how igb_uio would make any difference at all. If you are
using igb_uio, you *don't have DMA mapping at all* and will use raw
physical addresses. Assuming your code supports this, that's all you're
ever going to get. The point of VFIO is to have memory regions that are
mapped for DMA *because real physical addresses are assumed to be not
available*. When you're using igb_uio, you effectively do have DMA
access to the entire memory, and thus can bypass IOMMU altogether
(assuming you're using passthrough mode).
Bottom line: do VFIO DMA mapping unconditionally. If VFIO is active -
great, the memory will be DMA mapped. If it's not active - no harm will
ever be done by mapping the memory for DMA anyway.
--
Thanks,
Anatoly
