Return-Path: Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id S933379AbbDUPsf (ORCPT ); Tue, 21 Apr 2015 11:48:35 -0400 Received: from e06smtp14.uk.ibm.com ([195.75.94.110]:45821 "EHLO e06smtp14.uk.ibm.com" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S933163AbbDUPs3 (ORCPT ); Tue, 21 Apr 2015 11:48:29 -0400 Date: Tue, 21 Apr 2015 17:48:20 +0200 From: Greg Kurz To: "Michael S. Tsirkin" Cc: Rusty Russell , Cornelia Huck , linux-api@vger.kernel.org, linux-kernel@vger.kernel.org, kvm@vger.kernel.org, virtualization@lists.linux-foundation.org Subject: Re: [PATCH v4 7/8] vhost: feature to set the vring endianness Message-ID: <20150421174820.78c150b7@bahia.local> In-Reply-To: <20150421155753-mutt-send-email-mst@redhat.com> References: <20150410101500.31843.61248.stgit@bahia.local> <20150410101627.31843.21710.stgit@bahia.local> <20150421155753-mutt-send-email-mst@redhat.com> Organization: IBM X-Mailer: Claws Mail 3.11.1 (GTK+ 2.24.27; x86_64-redhat-linux-gnu) MIME-Version: 1.0 Content-Type: text/plain; charset=US-ASCII Content-Transfer-Encoding: 7bit X-TM-AS-MML: disable X-Content-Scanned: Fidelis XPS MAILER x-cbid: 15042115-0017-0000-0000-000003CB181E Sender: linux-kernel-owner@vger.kernel.org List-ID: X-Mailing-List: linux-kernel@vger.kernel.org Content-Length: 8642 Lines: 270 On Tue, 21 Apr 2015 16:04:23 +0200 "Michael S. Tsirkin" wrote: > On Fri, Apr 10, 2015 at 12:19:16PM +0200, Greg Kurz wrote: > > This patch brings cross-endian support to vhost when used to implement > > legacy virtio devices. Since it is a relatively rare situation, the > > feature availability is controlled by a kernel config option (not set > > by default). > > > > The vq->is_le boolean field is added to cache the endianness to be > > used for ring accesses. It defaults to native endian, as expected > > by legacy virtio devices. When the ring gets active, we force little > > endian if the device is modern. When the ring is deactivated, we > > revert to the native endian default. > > > > If cross-endian was compiled in, a vq->user_be boolean field is added > > so that userspace may request a specific endianness. This field is > > used to override the default when activating the ring of a legacy > > device. It has no effect on modern devices. > > > > Signed-off-by: Greg Kurz > > --- > > drivers/vhost/Kconfig | 10 ++++++the > > drivers/vhost/vhost.c | 76 +++++++++++++++++++++++++++++++++++++++++++- > > drivers/vhost/vhost.h | 12 +++++-- > > include/uapi/linux/vhost.h | 9 +++++ > > 4 files changed, 103 insertions(+), 4 deletions(-) > > > > Changes since v3: > > - VHOST_SET_VRING_ENDIAN_LEGACY ioctl renamed to VHOST_SET_VRING_BIG_ENDIAN > > - ioctl API is now: 0 for le, 1 for be, other values are EINVAL > > - ioctl doesn't filter out modern devices > > - ioctl stubs return ENOIOCTLCMD > > - forbid endianness changes when vring is active > > - logic now handled with vq->is_le and vq->user_be according to device > > start/stop as suggested by Michael > > > > diff --git a/drivers/vhost/Kconfig b/drivers/vhost/Kconfig > > index 017a1e8..0aec88c 100644 > > --- a/drivers/vhost/Kconfig > > +++ b/drivers/vhost/Kconfig > > @@ -32,3 +32,13 @@ config VHOST > > ---help--- > > This option is selected by any driver which needs to access > > the core of vhost. > > + > > +config VHOST_SET_ENDIAN_LEGACY > > I'd prefer namin this VHOST_CROSS_ENDIAN_LEGACY > Yes, makes sense. I'll make sure most of the cross-endian changes are easy to grep. > > + bool "Cross-endian support for host kernel accelerator" > > + default n > > + ---help--- > > + This option allows vhost to support guests with a different byte > > + ordering from host. It is disabled by default since it adds overhead > > + and it is only needed by a few platforms (powerpc and arm). > > and is only useful on a few platforms (powerpc and arm). > > "it" seems to refer to "overhead", which is rarely needed. > needed is a bit too strong, you can always e.g. run virtio > in userspace. > My poor English again... it seems you understood what I wanted to write though :) > > + > > + If unsure, say "N". > > diff --git a/drivers/vhost/vhost.c b/drivers/vhost/vhost.c > > index 2ee2826..3eb756b 100644the > > --- a/drivers/vhost/vhost.c > > +++ b/drivers/vhost/vhost.c > > @@ -199,6 +199,10 @@ static void vhost_vq_reset(struct vhost_dev *dev, > > vq->call = NULL; > > vq->log_ctx = NULL; > > vq->memory = NULL; > > + vq->is_le = virtio_legacy_is_little_endian(); > > +#ifdef CONFIG_VHOST_SET_ENDIAN_LEGACY > > + vq->user_be = !vq->is_le; > > +#endif > > add a wrapper for this too? > Will do. > > } > > > > static int vhost_worker(void *data) > > @@ -630,6 +634,53 @@ static long vhost_set_memory(struct vhost_dev *d, struct vhost_memory __user *m) > > return 0; > > } > > > > +#ifdef CONFIG_VHOST_SET_ENDIAN_LEGACY > > +static long vhost_set_vring_big_endian(struct vhost_virtqueue *vq, > > + int __user *argp) > > +{ > > + struct vhost_vring_state s; > > + > > + if (vq->private_data) > > + return -EBUSY; > > + > > + if (copy_from_user(&s, argp, sizeof(s))) > > + return -EFAULT; > > + > > + if (s.num && s.num != 1) > > s.num & ~0x1 > Since s.num is unsigned and I assume this won't change, what about s.num > 1 as suggested by Cornelia ? > > > + return -EINVAL; > > + > > + vq->user_be = s.num; > > + > > + return 0; > > +} > > + > > +static long vhost_get_vring_big_endian(struct vhost_virtqueue *vq, u32 idx, > > + int __user *argp) > > +{ > > + struct vhost_vring_state s = { > > + .index = idx, > > + .num = vq->user_be > > + }; > > + > > + if (copy_to_user(argp, &s, sizeof(s))) > > + return -EFAULT; > > + > > + return 0; > > +} > > +#else > > +static long vhost_set_vring_big_endian(struct vhost_virtqueue *vq, > > + int __user *argp) > > +{ > > + return -ENOIOCTLCMD; > > +} > > + > > +static long vhost_get_vring_big_endian(struct vhost_virtqueue *vq, u32 idx, > > + int __user *argp) > > +{ > > + return -ENOIOCTLCMD; > > +} > > +#endif /* CONFIG_VHOST_SET_ENDIAN_LEGACY */ > > + > > long vhost_vring_ioctl(struct vhost_dev *d, int ioctl, void __user *argp) > > { > > struct file *eventfp, *filep = NULL; > > @@ -806,6 +857,12 @@ long vhost_vring_ioctl(struct vhost_dev *d, int ioctl, void __user *argp) > > } else > > filep = eventfp;the > > break; > > + case VHOST_SET_VRING_BIG_ENDIAN: > > + r = vhost_set_vring_big_endian(vq, argp); > > + break; > > + case VHOST_GET_VRING_BIG_ENDIAN: > > + r = vhost_get_vring_big_endian(vq, idx, argp); > > + break; > > default: > > r = -ENOIOCTLCMD; > > } > > @@ -1040,12 +1097,29 @@ static int vhost_update_avail_event(struct vhost_virtqueue *vq, u16 avail_event) > > return 0; > > } > > > > +#ifdef CONFIG_VHOST_SET_ENDIAN_LEGACY > > +static void vhost_init_is_le(struct vhost_virtqueue *vq) > > +{ > > + vq->is_le = vhost_has_feature(vq, VIRTIO_F_VERSION_1) || !vq->user_be; > > +} > > +#else > > +static void vhost_init_is_le(struct vhost_virtqueue *vq) > > +{ > > + if (vhost_has_feature(vq, VIRTIO_F_VERSION_1)) > > + vq->is_le = true; > > +} > > +#endif > > + > > I'd prefer localizing ifdefery somewhere near top of file. > Will do. > > int vhost_init_used(struct vhost_virtqueue *vq) > > { > > __virtio16 last_used_idx; > > int r; > > - if (!vq->private_data) > > + if (!vq->private_data) { > > + vq->is_le = virtio_legacy_is_little_endian(); > > return 0; > > + } > > + > > + vhost_init_is_le(vq); > > > > r = vhost_update_used_flags(vq); > > if (r) > > diff --git a/drivers/vhost/vhost.h b/drivers/vhost/vhost.h > > index 4e9a186..04b2add 100644 > > --- a/drivers/vhost/vhost.h > > +++ b/drivers/vhost/vhost.h > > @@ -106,6 +106,14 @@ struct vhost_virtqueue { > > /* Log write descriptors */ > > void __user *log_base; > > struct vhost_log *log; > > + > > + /* Ring endianness. Defaults to legacy native endianness. > > + * Set to true when starting a modern virtio device. */ > > + bool is_le; > > +#ifdef CONFIG_VHOST_SET_ENDIAN_LEGACY > > + /* Ring endianness requested by userspace for cross-endian support. */ > > + bool user_be; > > +#endif > > }; > > > > struct vhost_dev { > > @@ -175,9 +183,7 @@ static inline bool vhost_has_feature(struct vhost_virtqueue *vq, int bit) > > > > static inline bool vhost_is_little_endian(struct vhost_virtqueue *vq) > > { > > - if (vhost_has_feature(vq, VIRTIO_F_VERSION_1)) > > - return true; > > - return virtio_legacy_is_little_endian(); > > + return vq->is_le; > > } > > > > /* Memory accessors */ > > diff --git a/include/uapi/linux/vhost.h b/include/uapi/linux/vhost.h > > index bb6a5b4..5cdebbc 100644 > > --- a/include/uapi/linux/vhost.h > > +++ b/include/uapi/linux/vhost.h > > @@ -103,6 +103,15 @@ struct vhost_memory { > > /* Get accessor: reads index, writes value in num */ > > #define VHOST_GET_VRING_BASE _IOWR(VHOST_VIRTIO, 0x12, struct vhost_vring_state) > > > > +/* Set the vring byte order in num. This is a legacy only API that is simply > > + * ignored when VIRTIO_F_VERSION_1 is set. > > + * 0 to set to little-endian > > + * 1 to set to big-endian > > How about defines for these? > Ok. I'll put the defines here so that all the cross-endian stuff lies in the same hunk. Is it ok for you ? > > + * other values return EINVAL. > > + */ > > +#define VHOST_SET_VRING_BIG_ENDIAN _IOW(VHOST_VIRTIO, 0x13, struct vhost_vring_state) > > +#define VHOST_GET_VRING_BIG_ENDIAN _IOW(VHOST_VIRTIO, 0x14, struct vhost_vring_state) > > + > > /* The following ioctls use eventfd file descriptors to signal and poll > > * for events. */ > > > -- To unsubscribe from this list: send the line "unsubscribe linux-kernel" in the body of a message to majordomo@vger.kernel.org More majordomo info at http://vger.kernel.org/majordomo-info.html Please read the FAQ at http://www.tux.org/lkml/