From: Stefano Stabellini <sstabellini@kernel.org> To: Greg Kurz <groug@kaod.org> Cc: Stefano Stabellini <sstabellini@kernel.org>, qemu-devel@nongnu.org, xen-devel@lists.xenproject.org, anthony.perard@citrix.com, Stefano Stabellini <stefano@aporeto.com>, jgross@suse.com Subject: Re: [Qemu-devel] [PATCH v2 3/9] xen: introduce the header file for the Xen 9pfs transport protocol Date: Wed, 15 Mar 2017 12:02:18 -0700 (PDT) [thread overview] Message-ID: <alpine.DEB.2.10.1703151136520.4807@sstabellini-ThinkPad-X260> (raw) In-Reply-To: <20170315100653.6702d52f@bahia.lab.toulouse-stg.fr.ibm.com> On Wed, 15 Mar 2017, Greg Kurz wrote: > On Mon, 13 Mar 2017 16:55:54 -0700 > Stefano Stabellini <sstabellini@kernel.org> wrote: > > > It uses the new ring.h macros to declare rings and interfaces. > > > > Signed-off-by: Stefano Stabellini <stefano@aporeto.com> > > CC: anthony.perard@citrix.com > > CC: jgross@suse.com > > --- > > hw/9pfs/xen_9pfs.h | 20 ++++++++++++++++++++ > > 1 file changed, 20 insertions(+) > > create mode 100644 hw/9pfs/xen_9pfs.h > > > > This header file has only one user: please move its content to > hw/9pfs/xen-9p-backend.c (except the 9P header structure, see > below). OK, I can do that. There is going to be a very similar header in the Xen tree under xen/include/public/io (http://marc.info/?l=xen-devel&m=148952997709142), but from QEMU point of view, it makes sense to fold it in xen-9p-backend.c. > > diff --git a/hw/9pfs/xen_9pfs.h b/hw/9pfs/xen_9pfs.h > > new file mode 100644 > > index 0000000..c4e8901 > > --- /dev/null > > +++ b/hw/9pfs/xen_9pfs.h > > @@ -0,0 +1,20 @@ > > +#ifndef XEN_9PFS_H > > +#define XEN_9PFS_H > > + > > +#include "hw/xen/io/ring.h" > > +#include <xen/io/protocols.h> > > + > > +struct xen_9pfs_header { > > + uint32_t size; > > + uint8_t id; > > + uint16_t tag; > > + > > + /* uint8_t sdata[]; */ > > This doesn't seem useful. I'll remove > > +} __attribute__((packed)); > > + > > A few remarks: > - this is a 9P message header actually, which is also used with virtio, > - QEMU coding style requires a typedef in CamelCase, > - the 9P protocol explicitely uses little-endian ordering. Since we > don't have endian-specific types, it makes sense to indicate that > when naming the fields. > - we have a QEMU_PACKED macro which seems to be used a lot more than > the gcc syntax > > Please define a new type in hw/9pfs/9p.h and use it in both transports. > Something like: > > typedef struct { > uint32_t size_le; > uint8_t id; > uint16_t tag_le; > } QEMU_PACKED P9MsgHeader; OK > > +#define PAGE_SHIFT XC_PAGE_SHIFT > > I don't see any user for this in hw/9pfs/xen-9p-backend.c PAGE_SHIFT is used by the macros below, but the original Xen headers don't have the PAGE_SHIFT definition, so, for the sake of keeping the two in sync, I didn't add it there. > > +#define XEN_9PFS_RING_ORDER 6 > > +#define XEN_9PFS_RING_SIZE XEN_FLEX_RING_SIZE(XEN_9PFS_RING_ORDER) > > +DEFINE_XEN_FLEX_RING_AND_INTF(xen_9pfs); > > + > > +#endif
WARNING: multiple messages have this Message-ID (diff)
From: Stefano Stabellini <sstabellini@kernel.org> To: Greg Kurz <groug@kaod.org> Cc: jgross@suse.com, Stefano Stabellini <sstabellini@kernel.org>, qemu-devel@nongnu.org, Stefano Stabellini <stefano@aporeto.com>, anthony.perard@citrix.com, xen-devel@lists.xenproject.org Subject: Re: [PATCH v2 3/9] xen: introduce the header file for the Xen 9pfs transport protocol Date: Wed, 15 Mar 2017 12:02:18 -0700 (PDT) [thread overview] Message-ID: <alpine.DEB.2.10.1703151136520.4807@sstabellini-ThinkPad-X260> (raw) In-Reply-To: <20170315100653.6702d52f@bahia.lab.toulouse-stg.fr.ibm.com> On Wed, 15 Mar 2017, Greg Kurz wrote: > On Mon, 13 Mar 2017 16:55:54 -0700 > Stefano Stabellini <sstabellini@kernel.org> wrote: > > > It uses the new ring.h macros to declare rings and interfaces. > > > > Signed-off-by: Stefano Stabellini <stefano@aporeto.com> > > CC: anthony.perard@citrix.com > > CC: jgross@suse.com > > --- > > hw/9pfs/xen_9pfs.h | 20 ++++++++++++++++++++ > > 1 file changed, 20 insertions(+) > > create mode 100644 hw/9pfs/xen_9pfs.h > > > > This header file has only one user: please move its content to > hw/9pfs/xen-9p-backend.c (except the 9P header structure, see > below). OK, I can do that. There is going to be a very similar header in the Xen tree under xen/include/public/io (http://marc.info/?l=xen-devel&m=148952997709142), but from QEMU point of view, it makes sense to fold it in xen-9p-backend.c. > > diff --git a/hw/9pfs/xen_9pfs.h b/hw/9pfs/xen_9pfs.h > > new file mode 100644 > > index 0000000..c4e8901 > > --- /dev/null > > +++ b/hw/9pfs/xen_9pfs.h > > @@ -0,0 +1,20 @@ > > +#ifndef XEN_9PFS_H > > +#define XEN_9PFS_H > > + > > +#include "hw/xen/io/ring.h" > > +#include <xen/io/protocols.h> > > + > > +struct xen_9pfs_header { > > + uint32_t size; > > + uint8_t id; > > + uint16_t tag; > > + > > + /* uint8_t sdata[]; */ > > This doesn't seem useful. I'll remove > > +} __attribute__((packed)); > > + > > A few remarks: > - this is a 9P message header actually, which is also used with virtio, > - QEMU coding style requires a typedef in CamelCase, > - the 9P protocol explicitely uses little-endian ordering. Since we > don't have endian-specific types, it makes sense to indicate that > when naming the fields. > - we have a QEMU_PACKED macro which seems to be used a lot more than > the gcc syntax > > Please define a new type in hw/9pfs/9p.h and use it in both transports. > Something like: > > typedef struct { > uint32_t size_le; > uint8_t id; > uint16_t tag_le; > } QEMU_PACKED P9MsgHeader; OK > > +#define PAGE_SHIFT XC_PAGE_SHIFT > > I don't see any user for this in hw/9pfs/xen-9p-backend.c PAGE_SHIFT is used by the macros below, but the original Xen headers don't have the PAGE_SHIFT definition, so, for the sake of keeping the two in sync, I didn't add it there. > > +#define XEN_9PFS_RING_ORDER 6 > > +#define XEN_9PFS_RING_SIZE XEN_FLEX_RING_SIZE(XEN_9PFS_RING_ORDER) > > +DEFINE_XEN_FLEX_RING_AND_INTF(xen_9pfs); > > + > > +#endif _______________________________________________ Xen-devel mailing list Xen-devel@lists.xen.org https://lists.xen.org/xen-devel
next prev parent reply other threads:[~2017-03-15 19:02 UTC|newest] Thread overview: 72+ messages / expand[flat|nested] mbox.gz Atom feed top 2017-03-13 23:55 [Qemu-devel] [PATCH v2 0/9] xen/9pfs: introduce the Xen 9pfs backend Stefano Stabellini 2017-03-13 23:55 ` Stefano Stabellini 2017-03-13 23:55 ` [Qemu-devel] [PATCH v2 1/9] configure: change CONFIG_XEN_BACKEND to be a target property Stefano Stabellini 2017-03-13 23:55 ` Stefano Stabellini 2017-03-13 23:55 ` [Qemu-devel] [PATCH v2 2/9] xen: import ring.h from xen Stefano Stabellini 2017-03-13 23:55 ` Stefano Stabellini 2017-03-15 8:00 ` [Qemu-devel] " Greg Kurz 2017-03-15 8:00 ` Greg Kurz 2017-03-15 18:36 ` [Qemu-devel] " Stefano Stabellini 2017-03-15 18:36 ` Stefano Stabellini 2017-03-15 19:00 ` [Qemu-devel] " Greg Kurz 2017-03-15 19:00 ` Greg Kurz 2017-03-15 19:04 ` [Qemu-devel] " Stefano Stabellini 2017-03-15 19:04 ` Stefano Stabellini 2017-03-13 23:55 ` [Qemu-devel] [PATCH v2 3/9] xen: introduce the header file for the Xen 9pfs transport protocol Stefano Stabellini 2017-03-13 23:55 ` Stefano Stabellini 2017-03-15 9:06 ` [Qemu-devel] " Greg Kurz 2017-03-15 9:06 ` Greg Kurz 2017-03-15 19:02 ` Stefano Stabellini [this message] 2017-03-15 19:02 ` Stefano Stabellini 2017-03-13 23:55 ` [Qemu-devel] [PATCH v2 4/9] xen/9pfs: introduce Xen 9pfs backend Stefano Stabellini 2017-03-13 23:55 ` Stefano Stabellini 2017-03-14 7:11 ` [Qemu-devel] " Juergen Gross 2017-03-14 7:11 ` Juergen Gross 2017-03-14 20:07 ` [Qemu-devel] " Stefano Stabellini 2017-03-14 20:07 ` Stefano Stabellini 2017-03-13 23:55 ` [Qemu-devel] [PATCH v2 5/9] xen/9pfs: connect to the frontend Stefano Stabellini 2017-03-13 23:55 ` Stefano Stabellini 2017-03-14 7:17 ` [Qemu-devel] " Juergen Gross 2017-03-14 7:17 ` Juergen Gross 2017-03-14 20:14 ` [Qemu-devel] " Stefano Stabellini 2017-03-14 20:14 ` Stefano Stabellini 2017-03-15 9:48 ` [Qemu-devel] " Greg Kurz 2017-03-15 9:48 ` Greg Kurz 2017-03-15 21:30 ` [Qemu-devel] " Stefano Stabellini 2017-03-15 21:30 ` Stefano Stabellini 2017-03-13 23:55 ` [Qemu-devel] [PATCH v2 6/9] xen/9pfs: receive requests from " Stefano Stabellini 2017-03-13 23:55 ` Stefano Stabellini 2017-03-15 10:51 ` [Qemu-devel] " Greg Kurz 2017-03-15 10:51 ` Greg Kurz 2017-03-15 21:56 ` [Qemu-devel] " Stefano Stabellini 2017-03-15 21:56 ` Stefano Stabellini 2017-03-15 11:53 ` Greg Kurz 2017-03-15 11:53 ` Greg Kurz 2017-03-15 22:03 ` [Qemu-devel] " Stefano Stabellini 2017-03-15 22:03 ` Stefano Stabellini 2017-03-13 23:55 ` [Qemu-devel] [PATCH v2 7/9] xen/9pfs: implement in/out_iov_from_pdu and vmarshal/vunmarshal Stefano Stabellini 2017-03-13 23:55 ` Stefano Stabellini 2017-03-15 10:58 ` [Qemu-devel] " Greg Kurz 2017-03-15 10:58 ` Greg Kurz 2017-03-15 21:58 ` [Qemu-devel] " Stefano Stabellini 2017-03-15 21:58 ` Stefano Stabellini 2017-03-13 23:55 ` [Qemu-devel] [PATCH v2 8/9] xen/9pfs: send responses back to the frontend Stefano Stabellini 2017-03-13 23:55 ` Stefano Stabellini 2017-03-15 10:59 ` [Qemu-devel] " Greg Kurz 2017-03-15 10:59 ` Greg Kurz 2017-03-15 21:59 ` [Qemu-devel] " Stefano Stabellini 2017-03-15 21:59 ` Stefano Stabellini 2017-03-13 23:56 ` [Qemu-devel] [PATCH v2 9/9] xen/9pfs: build and register Xen 9pfs backend Stefano Stabellini 2017-03-13 23:56 ` Stefano Stabellini 2017-03-14 10:02 ` [Qemu-devel] [PATCH v2 1/9] configure: change CONFIG_XEN_BACKEND to be a target property Paolo Bonzini 2017-03-14 10:02 ` Paolo Bonzini 2017-03-14 19:07 ` [Qemu-devel] " Stefano Stabellini 2017-03-14 19:07 ` Stefano Stabellini 2017-03-14 20:23 ` [Qemu-devel] " Stefano Stabellini 2017-03-14 20:23 ` Stefano Stabellini 2017-03-15 1:03 ` [Qemu-devel] " Greg Kurz 2017-03-15 1:03 ` Greg Kurz 2017-03-15 7:41 ` [Qemu-devel] " Paolo Bonzini 2017-03-15 7:41 ` Paolo Bonzini 2017-03-15 18:42 ` [Qemu-devel] " Stefano Stabellini 2017-03-15 18:42 ` Stefano Stabellini
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=alpine.DEB.2.10.1703151136520.4807@sstabellini-ThinkPad-X260 \ --to=sstabellini@kernel.org \ --cc=anthony.perard@citrix.com \ --cc=groug@kaod.org \ --cc=jgross@suse.com \ --cc=qemu-devel@nongnu.org \ --cc=stefano@aporeto.com \ --cc=xen-devel@lists.xenproject.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: linkBe 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.