From: Stefano Garzarella <sgarzare@redhat.com>
To: virtualization@lists.linux-foundation.org
Cc: stefanha@redhat.com, "Michael S. Tsirkin" <mst@redhat.com>,
Andrey Zhadchenko <andrey.zhadchenko@virtuozzo.com>,
eperezma@redhat.com, netdev@vger.kernel.org,
Jason Wang <jasowang@redhat.com>,
kvm@vger.kernel.org, linux-kernel@vger.kernel.org,
Stefano Garzarella <sgarzare@redhat.com>
Subject: [PATCH v3 4/8] vringh: support VA with iotlb
Date: Tue, 21 Mar 2023 16:42:24 +0100 [thread overview]
Message-ID: <20230321154228.182769-5-sgarzare@redhat.com> (raw)
In-Reply-To: <20230321154228.182769-1-sgarzare@redhat.com>
vDPA supports the possibility to use user VA in the iotlb messages.
So, let's add support for user VA in vringh to use it in the vDPA
simulators.
Signed-off-by: Stefano Garzarella <sgarzare@redhat.com>
---
Notes:
v3:
- refactored avoiding code duplication [Eugenio]
v2:
- replace kmap_atomic() with kmap_local_page() [see previous patch]
- fix cast warnings when build with W=1 C=1
include/linux/vringh.h | 5 +-
drivers/vdpa/mlx5/net/mlx5_vnet.c | 2 +-
drivers/vdpa/vdpa_sim/vdpa_sim.c | 4 +-
drivers/vhost/vringh.c | 153 +++++++++++++++++++++++-------
4 files changed, 127 insertions(+), 37 deletions(-)
diff --git a/include/linux/vringh.h b/include/linux/vringh.h
index 1991a02c6431..d39b9f2dcba0 100644
--- a/include/linux/vringh.h
+++ b/include/linux/vringh.h
@@ -32,6 +32,9 @@ struct vringh {
/* Can we get away with weak barriers? */
bool weak_barriers;
+ /* Use user's VA */
+ bool use_va;
+
/* Last available index we saw (ie. where we're up to). */
u16 last_avail_idx;
@@ -279,7 +282,7 @@ void vringh_set_iotlb(struct vringh *vrh, struct vhost_iotlb *iotlb,
spinlock_t *iotlb_lock);
int vringh_init_iotlb(struct vringh *vrh, u64 features,
- unsigned int num, bool weak_barriers,
+ unsigned int num, bool weak_barriers, bool use_va,
struct vring_desc *desc,
struct vring_avail *avail,
struct vring_used *used);
diff --git a/drivers/vdpa/mlx5/net/mlx5_vnet.c b/drivers/vdpa/mlx5/net/mlx5_vnet.c
index 520646ae7fa0..dfd0e000217b 100644
--- a/drivers/vdpa/mlx5/net/mlx5_vnet.c
+++ b/drivers/vdpa/mlx5/net/mlx5_vnet.c
@@ -2537,7 +2537,7 @@ static int setup_cvq_vring(struct mlx5_vdpa_dev *mvdev)
if (mvdev->actual_features & BIT_ULL(VIRTIO_NET_F_CTRL_VQ))
err = vringh_init_iotlb(&cvq->vring, mvdev->actual_features,
- MLX5_CVQ_MAX_ENT, false,
+ MLX5_CVQ_MAX_ENT, false, false,
(struct vring_desc *)(uintptr_t)cvq->desc_addr,
(struct vring_avail *)(uintptr_t)cvq->driver_addr,
(struct vring_used *)(uintptr_t)cvq->device_addr);
diff --git a/drivers/vdpa/vdpa_sim/vdpa_sim.c b/drivers/vdpa/vdpa_sim/vdpa_sim.c
index eea23c630f7c..47cdf2a1f5b8 100644
--- a/drivers/vdpa/vdpa_sim/vdpa_sim.c
+++ b/drivers/vdpa/vdpa_sim/vdpa_sim.c
@@ -60,7 +60,7 @@ static void vdpasim_queue_ready(struct vdpasim *vdpasim, unsigned int idx)
struct vdpasim_virtqueue *vq = &vdpasim->vqs[idx];
uint16_t last_avail_idx = vq->vring.last_avail_idx;
- vringh_init_iotlb(&vq->vring, vdpasim->features, vq->num, true,
+ vringh_init_iotlb(&vq->vring, vdpasim->features, vq->num, true, false,
(struct vring_desc *)(uintptr_t)vq->desc_addr,
(struct vring_avail *)
(uintptr_t)vq->driver_addr,
@@ -92,7 +92,7 @@ static void vdpasim_vq_reset(struct vdpasim *vdpasim,
vq->cb = NULL;
vq->private = NULL;
vringh_init_iotlb(&vq->vring, vdpasim->dev_attr.supported_features,
- VDPASIM_QUEUE_MAX, false, NULL, NULL, NULL);
+ VDPASIM_QUEUE_MAX, false, false, NULL, NULL, NULL);
vq->vring.notify = NULL;
}
diff --git a/drivers/vhost/vringh.c b/drivers/vhost/vringh.c
index 0ba3ef809e48..72c88519329a 100644
--- a/drivers/vhost/vringh.c
+++ b/drivers/vhost/vringh.c
@@ -1094,10 +1094,18 @@ EXPORT_SYMBOL(vringh_need_notify_kern);
#if IS_REACHABLE(CONFIG_VHOST_IOTLB)
+struct iotlb_vec {
+ union {
+ struct iovec *iovec;
+ struct bio_vec *bvec;
+ } iov;
+ size_t count;
+ bool is_iovec;
+};
+
static int iotlb_translate(const struct vringh *vrh,
u64 addr, u64 len, u64 *translated,
- struct bio_vec iov[],
- int iov_size, u32 perm)
+ struct iotlb_vec *ivec, u32 perm)
{
struct vhost_iotlb_map *map;
struct vhost_iotlb *iotlb = vrh->iotlb;
@@ -1107,9 +1115,9 @@ static int iotlb_translate(const struct vringh *vrh,
spin_lock(vrh->iotlb_lock);
while (len > s) {
- u64 size, pa, pfn;
+ u64 size;
- if (unlikely(ret >= iov_size)) {
+ if (unlikely(ret >= ivec->count)) {
ret = -ENOBUFS;
break;
}
@@ -1124,10 +1132,22 @@ static int iotlb_translate(const struct vringh *vrh,
}
size = map->size - addr + map->start;
- pa = map->addr + addr - map->start;
- pfn = pa >> PAGE_SHIFT;
- bvec_set_page(&iov[ret], pfn_to_page(pfn), min(len - s, size),
- pa & (PAGE_SIZE - 1));
+ if (ivec->is_iovec) {
+ struct iovec *iovec = ivec->iov.iovec;
+
+ iovec[ret].iov_len = min(len - s, size);
+ iovec[ret].iov_base = (void __user *)(unsigned long)
+ (map->addr + addr - map->start);
+ } else {
+ u64 pa = map->addr + addr - map->start;
+ u64 pfn = pa >> PAGE_SHIFT;
+ struct bio_vec *bvec = ivec->iov.bvec;
+
+ bvec_set_page(&bvec[ret], pfn_to_page(pfn),
+ min(len - s, size),
+ pa & (PAGE_SIZE - 1));
+ }
+
s += size;
addr += size;
++ret;
@@ -1141,26 +1161,42 @@ static int iotlb_translate(const struct vringh *vrh,
return ret;
}
+#define IOTLB_IOV_SIZE 16
+
static inline int copy_from_iotlb(const struct vringh *vrh, void *dst,
void *src, size_t len)
{
+ struct iotlb_vec ivec;
+ union {
+ struct iovec iovec[IOTLB_IOV_SIZE];
+ struct bio_vec bvec[IOTLB_IOV_SIZE];
+ } iov;
u64 total_translated = 0;
+ ivec.iov.iovec = iov.iovec;
+ ivec.count = IOTLB_IOV_SIZE;
+ ivec.is_iovec = vrh->use_va;
+
while (total_translated < len) {
- struct bio_vec iov[16];
struct iov_iter iter;
u64 translated;
int ret;
ret = iotlb_translate(vrh, (u64)(uintptr_t)src,
len - total_translated, &translated,
- iov, ARRAY_SIZE(iov), VHOST_MAP_RO);
+ &ivec, VHOST_MAP_RO);
if (ret == -ENOBUFS)
- ret = ARRAY_SIZE(iov);
+ ret = IOTLB_IOV_SIZE;
else if (ret < 0)
return ret;
- iov_iter_bvec(&iter, ITER_SOURCE, iov, ret, translated);
+ if (ivec.is_iovec) {
+ iov_iter_init(&iter, ITER_SOURCE, ivec.iov.iovec, ret,
+ translated);
+ } else {
+ iov_iter_bvec(&iter, ITER_SOURCE, ivec.iov.bvec, ret,
+ translated);
+ }
ret = copy_from_iter(dst, translated, &iter);
if (ret < 0)
@@ -1177,23 +1213,37 @@ static inline int copy_from_iotlb(const struct vringh *vrh, void *dst,
static inline int copy_to_iotlb(const struct vringh *vrh, void *dst,
void *src, size_t len)
{
+ struct iotlb_vec ivec;
+ union {
+ struct iovec iovec[IOTLB_IOV_SIZE];
+ struct bio_vec bvec[IOTLB_IOV_SIZE];
+ } iov;
u64 total_translated = 0;
+ ivec.iov.iovec = iov.iovec;
+ ivec.count = IOTLB_IOV_SIZE;
+ ivec.is_iovec = vrh->use_va;
+
while (total_translated < len) {
- struct bio_vec iov[16];
struct iov_iter iter;
u64 translated;
int ret;
ret = iotlb_translate(vrh, (u64)(uintptr_t)dst,
len - total_translated, &translated,
- iov, ARRAY_SIZE(iov), VHOST_MAP_WO);
+ &ivec, VHOST_MAP_WO);
if (ret == -ENOBUFS)
- ret = ARRAY_SIZE(iov);
+ ret = IOTLB_IOV_SIZE;
else if (ret < 0)
return ret;
- iov_iter_bvec(&iter, ITER_DEST, iov, ret, translated);
+ if (ivec.is_iovec) {
+ iov_iter_init(&iter, ITER_DEST, ivec.iov.iovec, ret,
+ translated);
+ } else {
+ iov_iter_bvec(&iter, ITER_DEST, ivec.iov.bvec, ret,
+ translated);
+ }
ret = copy_to_iter(src, translated, &iter);
if (ret < 0)
@@ -1210,20 +1260,37 @@ static inline int copy_to_iotlb(const struct vringh *vrh, void *dst,
static inline int getu16_iotlb(const struct vringh *vrh,
u16 *val, const __virtio16 *p)
{
- struct bio_vec iov;
- void *kaddr, *from;
+ struct iotlb_vec ivec;
+ union {
+ struct iovec iovec[1];
+ struct bio_vec bvec[1];
+ } iov;
+ __virtio16 tmp;
int ret;
+ ivec.iov.iovec = iov.iovec;
+ ivec.count = 1;
+ ivec.is_iovec = vrh->use_va;
+
/* Atomic read is needed for getu16 */
- ret = iotlb_translate(vrh, (u64)(uintptr_t)p, sizeof(*p), NULL,
- &iov, 1, VHOST_MAP_RO);
+ ret = iotlb_translate(vrh, (u64)(uintptr_t)p, sizeof(*p),
+ NULL, &ivec, VHOST_MAP_RO);
if (ret < 0)
return ret;
- kaddr = kmap_local_page(iov.bv_page);
- from = kaddr + iov.bv_offset;
- *val = vringh16_to_cpu(vrh, READ_ONCE(*(__virtio16 *)from));
- kunmap_local(kaddr);
+ if (ivec.is_iovec) {
+ ret = __get_user(tmp, (__virtio16 __user *)ivec.iov.iovec[0].iov_base);
+ if (ret)
+ return ret;
+ } else {
+ void *kaddr = kmap_local_page(ivec.iov.bvec[0].bv_page);
+ void *from = kaddr + ivec.iov.bvec[0].bv_offset;
+
+ tmp = READ_ONCE(*(__virtio16 *)from);
+ kunmap_local(kaddr);
+ }
+
+ *val = vringh16_to_cpu(vrh, tmp);
return 0;
}
@@ -1231,20 +1298,37 @@ static inline int getu16_iotlb(const struct vringh *vrh,
static inline int putu16_iotlb(const struct vringh *vrh,
__virtio16 *p, u16 val)
{
- struct bio_vec iov;
- void *kaddr, *to;
+ struct iotlb_vec ivec;
+ union {
+ struct iovec iovec;
+ struct bio_vec bvec;
+ } iov;
+ __virtio16 tmp;
int ret;
+ ivec.iov.iovec = &iov.iovec;
+ ivec.count = 1;
+ ivec.is_iovec = vrh->use_va;
+
/* Atomic write is needed for putu16 */
- ret = iotlb_translate(vrh, (u64)(uintptr_t)p, sizeof(*p), NULL,
- &iov, 1, VHOST_MAP_WO);
+ ret = iotlb_translate(vrh, (u64)(uintptr_t)p, sizeof(*p),
+ NULL, &ivec, VHOST_MAP_RO);
if (ret < 0)
return ret;
- kaddr = kmap_local_page(iov.bv_page);
- to = kaddr + iov.bv_offset;
- WRITE_ONCE(*(__virtio16 *)to, cpu_to_vringh16(vrh, val));
- kunmap_local(kaddr);
+ tmp = cpu_to_vringh16(vrh, val);
+
+ if (ivec.is_iovec) {
+ ret = __put_user(tmp, (__virtio16 __user *)ivec.iov.iovec[0].iov_base);
+ if (ret)
+ return ret;
+ } else {
+ void *kaddr = kmap_local_page(ivec.iov.bvec[0].bv_page);
+ void *to = kaddr + ivec.iov.bvec[0].bv_offset;
+
+ WRITE_ONCE(*(__virtio16 *)to, tmp);
+ kunmap_local(kaddr);
+ }
return 0;
}
@@ -1306,6 +1390,7 @@ static inline int putused_iotlb(const struct vringh *vrh,
* @features: the feature bits for this ring.
* @num: the number of elements.
* @weak_barriers: true if we only need memory barriers, not I/O.
+ * @use_va: true if IOTLB contains user VA
* @desc: the userpace descriptor pointer.
* @avail: the userpace avail pointer.
* @used: the userpace used pointer.
@@ -1313,11 +1398,13 @@ static inline int putused_iotlb(const struct vringh *vrh,
* Returns an error if num is invalid.
*/
int vringh_init_iotlb(struct vringh *vrh, u64 features,
- unsigned int num, bool weak_barriers,
+ unsigned int num, bool weak_barriers, bool use_va,
struct vring_desc *desc,
struct vring_avail *avail,
struct vring_used *used)
{
+ vrh->use_va = use_va;
+
return vringh_init_kern(vrh, features, num, weak_barriers,
desc, avail, used);
}
--
2.39.2
next prev parent reply other threads:[~2023-03-21 15:44 UTC|newest]
Thread overview: 28+ messages / expand[flat|nested] mbox.gz Atom feed top
2023-03-21 15:42 [PATCH v3 0/8] vdpa_sim: add support for user VA Stefano Garzarella
2023-03-21 15:42 ` [PATCH v3 1/8] vdpa: add bind_mm/unbind_mm callbacks Stefano Garzarella
2023-03-23 2:58 ` Jason Wang
2023-03-21 15:42 ` [PATCH v3 2/8] vhost-vdpa: use bind_mm/unbind_mm device callbacks Stefano Garzarella
2023-03-23 3:01 ` Jason Wang
2023-03-23 9:38 ` Stefano Garzarella
2023-03-21 15:42 ` [PATCH v3 3/8] vringh: replace kmap_atomic() with kmap_local_page() Stefano Garzarella
2023-03-22 10:37 ` Fabio M. De Francesco
2023-03-23 3:02 ` Jason Wang
2023-03-21 15:42 ` Stefano Garzarella [this message]
2023-03-23 3:36 ` [PATCH v3 4/8] vringh: support VA with iotlb Jason Wang
2023-03-23 10:37 ` Stefano Garzarella
2023-03-23 8:09 ` Eugenio Perez Martin
2023-03-23 10:46 ` Stefano Garzarella
2023-03-23 14:43 ` Eugenio Perez Martin
2023-03-24 14:39 ` Stefano Garzarella
2023-03-21 15:48 ` [PATCH v3 5/8] vdpa_sim: make devices agnostic for work management Stefano Garzarella
2023-03-21 15:48 ` [PATCH v3 6/8] vdpa_sim: use kthread worker Stefano Garzarella
2023-03-21 15:48 ` [PATCH v3 7/8] vdpa_sim: replace the spinlock with a mutex to protect the state Stefano Garzarella
2023-03-21 15:48 ` [PATCH v3 8/8] vdpa_sim: add support for user VA Stefano Garzarella
2023-03-23 3:42 ` Jason Wang
2023-03-23 9:50 ` Stefano Garzarella
2023-03-23 11:44 ` Michael S. Tsirkin
2023-03-24 2:54 ` Jason Wang
2023-03-24 14:43 ` Stefano Garzarella
2023-03-27 3:12 ` Jason Wang
2023-03-24 3:49 ` Jason Wang
2023-03-24 14:46 ` Stefano Garzarella
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=20230321154228.182769-5-sgarzare@redhat.com \
--to=sgarzare@redhat.com \
--cc=andrey.zhadchenko@virtuozzo.com \
--cc=eperezma@redhat.com \
--cc=jasowang@redhat.com \
--cc=kvm@vger.kernel.org \
--cc=linux-kernel@vger.kernel.org \
--cc=mst@redhat.com \
--cc=netdev@vger.kernel.org \
--cc=stefanha@redhat.com \
--cc=virtualization@lists.linux-foundation.org \
/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).