* [For Linux 0/4] PV protocol headers for Linux Kernel
@ 2017-04-07 8:30 Oleksandr Andrushchenko
2017-04-07 8:30 ` [For Linux 1/4] xen/kbdif: update protocol description Oleksandr Andrushchenko
` (9 more replies)
0 siblings, 10 replies; 41+ messages in thread
From: Oleksandr Andrushchenko @ 2017-04-07 8:30 UTC (permalink / raw)
To: xen-devel
Cc: joculator, al1img, vlad.babchuk, andrii.anisov, olekstysh,
konrad.wilk, lars.kurth, julien.grall, andr2000, sstabellini,
linux-kernel, Oleksandr Andrushchenko
From: Oleksandr Andrushchenko <oleksandr_andrushchenko@epam.com>
Hi, all!
This patch series adds/updates para-virtual device
protocols for Linux Kernel (headers):
o kbdif (multitouch support added)
o sndif - sound (new)
o displif - display (new)
No changes, but Linux Kernel coding style applied.
Thank you,
Oleksandr
Oleksandr Andrushchenko (4):
xen/kbdif: update protocol description
xen/kbdif: add multi-touch support
xen/sndif: add sound-device ABI
xen/displif: add ABI for para-virtual display
include/xen/interface/io/displif.h | 864 +++++++++++++++++++++++++++++++++++++
include/xen/interface/io/kbdif.h | 468 ++++++++++++++++++--
include/xen/interface/io/sndif.h | 803 ++++++++++++++++++++++++++++++++++
3 files changed, 2108 insertions(+), 27 deletions(-)
create mode 100644 include/xen/interface/io/displif.h
create mode 100644 include/xen/interface/io/sndif.h
--
2.7.4
^ permalink raw reply [flat|nested] 41+ messages in thread
* [For Linux 1/4] xen/kbdif: update protocol description
2017-04-07 8:30 [For Linux 0/4] PV protocol headers for Linux Kernel Oleksandr Andrushchenko
2017-04-07 8:30 ` [For Linux 1/4] xen/kbdif: update protocol description Oleksandr Andrushchenko
@ 2017-04-07 8:30 ` Oleksandr Andrushchenko
2017-04-07 11:29 ` [Xen-devel] " Juergen Gross
2017-04-07 11:29 ` Juergen Gross
2017-04-07 8:30 ` [For Linux 2/4] xen/kbdif: add multi-touch support Oleksandr Andrushchenko
` (7 subsequent siblings)
9 siblings, 2 replies; 41+ messages in thread
From: Oleksandr Andrushchenko @ 2017-04-07 8:30 UTC (permalink / raw)
To: xen-devel
Cc: joculator, al1img, vlad.babchuk, andrii.anisov, olekstysh,
konrad.wilk, lars.kurth, julien.grall, andr2000, sstabellini,
linux-kernel, Oleksandr Andrushchenko
From: Oleksandr Andrushchenko <oleksandr_andrushchenko@epam.com>
The patch clarifies the protocol that is used by the PV keyboard
drivers.
Reviewed-by: Konrad Rzeszutek Wilk <konrad.wilk@oracle.com>
Reviewed-by: Stefano Stabellini <sstabellini@kernel.org>
Signed-off-by: Oleksandr Andrushchenko <oleksandr_andrushchenko@epam.com>
---
include/xen/interface/io/kbdif.h | 258 +++++++++++++++++++++++++++++++++++----
1 file changed, 231 insertions(+), 27 deletions(-)
diff --git a/include/xen/interface/io/kbdif.h b/include/xen/interface/io/kbdif.h
index 8066c7849fbe..e83eaddb20dc 100644
--- a/include/xen/interface/io/kbdif.h
+++ b/include/xen/interface/io/kbdif.h
@@ -26,43 +26,223 @@
#ifndef __XEN_PUBLIC_IO_KBDIF_H__
#define __XEN_PUBLIC_IO_KBDIF_H__
-/* In events (backend -> frontend) */
+/*
+ *****************************************************************************
+ * Feature and Parameter Negotiation
+ *****************************************************************************
+ *
+ * The two halves of a para-virtual driver utilize nodes within
+ * XenStore to communicate capabilities and to negotiate operating parameters.
+ * This section enumerates these nodes which reside in the respective front and
+ * backend portions of XenStore, following XenBus convention.
+ *
+ * All data in XenStore is stored as strings. Nodes specifying numeric
+ * values are encoded in decimal. Integer value ranges listed below are
+ * expressed as fixed sized integer types capable of storing the conversion
+ * of a properly formated node string, without loss of information.
+ *
+ *****************************************************************************
+ * Backend XenBus Nodes
+ *****************************************************************************
+ *
+ *---------------------------- Features supported ----------------------------
+ *
+ * Capable backend advertises supported features by publishing
+ * corresponding entries in XenStore and puts 1 as the value of the entry.
+ * If a feature is not supported then 0 must be set or feature entry omitted.
+ *
+ * feature-abs-pointer
+ * Values: <uint>
+ *
+ * Backends, which support reporting of absolute coordinates for pointer
+ * device should set this to 1.
+ *
+ *------------------------- Pointer Device Parameters ------------------------
+ *
+ * width
+ * Values: <uint>
+ *
+ * Maximum X coordinate (width) to be used by the frontend
+ * while reporting input events, pixels, [0; UINT32_MAX].
+ *
+ * height
+ * Values: <uint>
+ *
+ * Maximum Y coordinate (height) to be used by the frontend
+ * while reporting input events, pixels, [0; UINT32_MAX].
+ *
+ *****************************************************************************
+ * Frontend XenBus Nodes
+ *****************************************************************************
+ *
+ *------------------------------ Feature request -----------------------------
+ *
+ * Capable frontend requests features from backend via setting corresponding
+ * entries to 1 in XenStore. Requests for features not advertised as supported
+ * by the backend have no effect.
+ *
+ * request-abs-pointer
+ * Values: <uint>
+ *
+ * Request backend to report absolute pointer coordinates
+ * (XENKBD_TYPE_POS) instead of relative ones (XENKBD_TYPE_MOTION).
+ *
+ *----------------------- Request Transport Parameters -----------------------
+ *
+ * event-channel
+ * Values: <uint>
+ *
+ * The identifier of the Xen event channel used to signal activity
+ * in the ring buffer.
+ *
+ * page-gref
+ * Values: <uint>
+ *
+ * The Xen grant reference granting permission for the backend to map
+ * a sole page in a single page sized event ring buffer.
+ *
+ * page-ref
+ * Values: <uint>
+ *
+ * OBSOLETE, not recommended for use.
+ * PFN of the shared page.
+ */
/*
- * Frontends should ignore unknown in events.
+ * EVENT CODES.
+ */
+
+#define XENKBD_TYPE_MOTION 1
+#define XENKBD_TYPE_RESERVED 2
+#define XENKBD_TYPE_KEY 3
+#define XENKBD_TYPE_POS 4
+
+/*
+ * CONSTANTS, XENSTORE FIELD AND PATH NAME STRINGS, HELPERS.
*/
-/* Pointer movement event */
-#define XENKBD_TYPE_MOTION 1
-/* Event type 2 currently not used */
-/* Key event (includes pointer buttons) */
-#define XENKBD_TYPE_KEY 3
+#define XENKBD_DRIVER_NAME "vkbd"
+
+#define XENKBD_FIELD_FEAT_ABS_POINTER "feature-abs-pointer"
+#define XENKBD_FIELD_REQ_ABS_POINTER "request-abs-pointer"
+#define XENKBD_FIELD_RING_GREF "page-gref"
+#define XENKBD_FIELD_EVT_CHANNEL "event-channel"
+#define XENKBD_FIELD_WIDTH "width"
+#define XENKBD_FIELD_HEIGHT "height"
+
+/* OBSOLETE, not recommended for use */
+#define XENKBD_FIELD_RING_REF "page-ref"
+
/*
- * Pointer position event
- * Capable backend sets feature-abs-pointer in xenstore.
- * Frontend requests ot instead of XENKBD_TYPE_MOTION by setting
- * request-abs-update in xenstore.
+ *****************************************************************************
+ * Description of the protocol between frontend and backend driver.
+ *****************************************************************************
+ *
+ * The two halves of a Para-virtual driver communicate with
+ * each other using a shared page and an event channel.
+ * Shared page contains a ring with event structures.
+ *
+ * All reserved fields in the structures below must be 0.
+ *
+ *****************************************************************************
+ * Backend to frontend events
+ *****************************************************************************
+ *
+ * Frontends should ignore unknown in events.
+ * All event packets have the same length (40 octets)
+ * All event packets have common header:
+ *
+ * 0 octet
+ * +-----------------+
+ * | type |
+ * +-----------------+
+ * type - uint8_t, event code, XENKBD_TYPE_???
+ *
+ *
+ * Pointer relative movement event
+ * 0 1 2 3 octet
+ * +----------------+----------------+----------------+----------------+
+ * | _TYPE_MOTION | reserved | 4
+ * +----------------+----------------+----------------+----------------+
+ * | rel_x | 8
+ * +----------------+----------------+----------------+----------------+
+ * | rel_y | 12
+ * +----------------+----------------+----------------+----------------+
+ * | rel_z | 16
+ * +----------------+----------------+----------------+----------------+
+ * | reserved | 20
+ * +----------------+----------------+----------------+----------------+
+ * |/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/|
+ * +----------------+----------------+----------------+----------------+
+ * | reserved | 40
+ * +----------------+----------------+----------------+----------------+
+ *
+ * rel_x - int32_t, relative X motion
+ * rel_y - int32_t, relative Y motion
+ * rel_z - int32_t, relative Z motion (wheel)
*/
-#define XENKBD_TYPE_POS 4
struct xenkbd_motion {
- uint8_t type; /* XENKBD_TYPE_MOTION */
- int32_t rel_x; /* relative X motion */
- int32_t rel_y; /* relative Y motion */
- int32_t rel_z; /* relative Z motion (wheel) */
+ uint8_t type;
+ int32_t rel_x;
+ int32_t rel_y;
+ int32_t rel_z;
};
+/*
+ * Key event (includes pointer buttons)
+ * 0 1 2 3 octet
+ * +----------------+----------------+----------------+----------------+
+ * | _TYPE_KEY | pressed | reserved | 4
+ * +----------------+----------------+----------------+----------------+
+ * | keycode | 8
+ * +----------------+----------------+----------------+----------------+
+ * | reserved | 12
+ * +----------------+----------------+----------------+----------------+
+ * |/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/|
+ * +----------------+----------------+----------------+----------------+
+ * | reserved | 40
+ * +----------------+----------------+----------------+----------------+
+ *
+ * pressed - uint8_t, 1 if pressed; 0 otherwise
+ * keycode - uint32_t, KEY_* from linux/input.h
+ */
+
struct xenkbd_key {
- uint8_t type; /* XENKBD_TYPE_KEY */
- uint8_t pressed; /* 1 if pressed; 0 otherwise */
- uint32_t keycode; /* KEY_* from linux/input.h */
+ uint8_t type;
+ uint8_t pressed;
+ uint32_t keycode;
};
+/*
+ * Pointer absolute position event
+ * 0 1 2 3 octet
+ * +----------------+----------------+----------------+----------------+
+ * | _TYPE_POS | reserved | 4
+ * +----------------+----------------+----------------+----------------+
+ * | abs_x | 8
+ * +----------------+----------------+----------------+----------------+
+ * | abs_y | 12
+ * +----------------+----------------+----------------+----------------+
+ * | rel_z | 16
+ * +----------------+----------------+----------------+----------------+
+ * | reserved | 20
+ * +----------------+----------------+----------------+----------------+
+ * |/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/|
+ * +----------------+----------------+----------------+----------------+
+ * | reserved | 40
+ * +----------------+----------------+----------------+----------------+
+ *
+ * abs_x - int32_t, absolute X position (in FB pixels)
+ * abs_y - int32_t, absolute Y position (in FB pixels)
+ * rel_z - int32_t, relative Z motion (wheel)
+ */
+
struct xenkbd_position {
- uint8_t type; /* XENKBD_TYPE_POS */
- int32_t abs_x; /* absolute X position (in FB pixels) */
- int32_t abs_y; /* absolute Y position (in FB pixels) */
- int32_t rel_z; /* relative Z motion (wheel) */
+ uint8_t type;
+ int32_t abs_x;
+ int32_t abs_y;
+ int32_t rel_z;
};
#define XENKBD_IN_EVENT_SIZE 40
@@ -75,12 +255,22 @@ union xenkbd_in_event {
char pad[XENKBD_IN_EVENT_SIZE];
};
-/* Out events (frontend -> backend) */
-
/*
+ *****************************************************************************
+ * Frontend to backend events
+ *****************************************************************************
+ *
* Out events may be sent only when requested by backend, and receipt
* of an unknown out event is an error.
* No out events currently defined.
+
+ * All event packets have the same length (40 octets)
+ * All event packets have common header:
+ * 0 octet
+ * +-----------------+
+ * | type |
+ * +-----------------+
+ * type - uint8_t, event code
*/
#define XENKBD_OUT_EVENT_SIZE 40
@@ -90,7 +280,11 @@ union xenkbd_out_event {
char pad[XENKBD_OUT_EVENT_SIZE];
};
-/* shared page */
+/*
+ *****************************************************************************
+ * Shared page
+ *****************************************************************************
+ */
#define XENKBD_IN_RING_SIZE 2048
#define XENKBD_IN_RING_LEN (XENKBD_IN_RING_SIZE / XENKBD_IN_EVENT_SIZE)
@@ -113,4 +307,14 @@ struct xenkbd_page {
uint32_t out_cons, out_prod;
};
-#endif
+#endif /* __XEN_PUBLIC_IO_KBDIF_H__ */
+
+/*
+ * Local variables:
+ * mode: C
+ * c-file-style: "BSD"
+ * c-basic-offset: 4
+ * tab-width: 4
+ * indent-tabs-mode: nil
+ * End:
+ */
--
2.7.4
^ permalink raw reply related [flat|nested] 41+ messages in thread
* [For Linux 1/4] xen/kbdif: update protocol description
2017-04-07 8:30 [For Linux 0/4] PV protocol headers for Linux Kernel Oleksandr Andrushchenko
@ 2017-04-07 8:30 ` Oleksandr Andrushchenko
2017-04-07 8:30 ` Oleksandr Andrushchenko
` (8 subsequent siblings)
9 siblings, 0 replies; 41+ messages in thread
From: Oleksandr Andrushchenko @ 2017-04-07 8:30 UTC (permalink / raw)
To: xen-devel
Cc: lars.kurth, sstabellini, vlad.babchuk, linux-kernel,
Oleksandr Andrushchenko, julien.grall, andrii.anisov, olekstysh,
andr2000, al1img, joculator
From: Oleksandr Andrushchenko <oleksandr_andrushchenko@epam.com>
The patch clarifies the protocol that is used by the PV keyboard
drivers.
Reviewed-by: Konrad Rzeszutek Wilk <konrad.wilk@oracle.com>
Reviewed-by: Stefano Stabellini <sstabellini@kernel.org>
Signed-off-by: Oleksandr Andrushchenko <oleksandr_andrushchenko@epam.com>
---
include/xen/interface/io/kbdif.h | 258 +++++++++++++++++++++++++++++++++++----
1 file changed, 231 insertions(+), 27 deletions(-)
diff --git a/include/xen/interface/io/kbdif.h b/include/xen/interface/io/kbdif.h
index 8066c7849fbe..e83eaddb20dc 100644
--- a/include/xen/interface/io/kbdif.h
+++ b/include/xen/interface/io/kbdif.h
@@ -26,43 +26,223 @@
#ifndef __XEN_PUBLIC_IO_KBDIF_H__
#define __XEN_PUBLIC_IO_KBDIF_H__
-/* In events (backend -> frontend) */
+/*
+ *****************************************************************************
+ * Feature and Parameter Negotiation
+ *****************************************************************************
+ *
+ * The two halves of a para-virtual driver utilize nodes within
+ * XenStore to communicate capabilities and to negotiate operating parameters.
+ * This section enumerates these nodes which reside in the respective front and
+ * backend portions of XenStore, following XenBus convention.
+ *
+ * All data in XenStore is stored as strings. Nodes specifying numeric
+ * values are encoded in decimal. Integer value ranges listed below are
+ * expressed as fixed sized integer types capable of storing the conversion
+ * of a properly formated node string, without loss of information.
+ *
+ *****************************************************************************
+ * Backend XenBus Nodes
+ *****************************************************************************
+ *
+ *---------------------------- Features supported ----------------------------
+ *
+ * Capable backend advertises supported features by publishing
+ * corresponding entries in XenStore and puts 1 as the value of the entry.
+ * If a feature is not supported then 0 must be set or feature entry omitted.
+ *
+ * feature-abs-pointer
+ * Values: <uint>
+ *
+ * Backends, which support reporting of absolute coordinates for pointer
+ * device should set this to 1.
+ *
+ *------------------------- Pointer Device Parameters ------------------------
+ *
+ * width
+ * Values: <uint>
+ *
+ * Maximum X coordinate (width) to be used by the frontend
+ * while reporting input events, pixels, [0; UINT32_MAX].
+ *
+ * height
+ * Values: <uint>
+ *
+ * Maximum Y coordinate (height) to be used by the frontend
+ * while reporting input events, pixels, [0; UINT32_MAX].
+ *
+ *****************************************************************************
+ * Frontend XenBus Nodes
+ *****************************************************************************
+ *
+ *------------------------------ Feature request -----------------------------
+ *
+ * Capable frontend requests features from backend via setting corresponding
+ * entries to 1 in XenStore. Requests for features not advertised as supported
+ * by the backend have no effect.
+ *
+ * request-abs-pointer
+ * Values: <uint>
+ *
+ * Request backend to report absolute pointer coordinates
+ * (XENKBD_TYPE_POS) instead of relative ones (XENKBD_TYPE_MOTION).
+ *
+ *----------------------- Request Transport Parameters -----------------------
+ *
+ * event-channel
+ * Values: <uint>
+ *
+ * The identifier of the Xen event channel used to signal activity
+ * in the ring buffer.
+ *
+ * page-gref
+ * Values: <uint>
+ *
+ * The Xen grant reference granting permission for the backend to map
+ * a sole page in a single page sized event ring buffer.
+ *
+ * page-ref
+ * Values: <uint>
+ *
+ * OBSOLETE, not recommended for use.
+ * PFN of the shared page.
+ */
/*
- * Frontends should ignore unknown in events.
+ * EVENT CODES.
+ */
+
+#define XENKBD_TYPE_MOTION 1
+#define XENKBD_TYPE_RESERVED 2
+#define XENKBD_TYPE_KEY 3
+#define XENKBD_TYPE_POS 4
+
+/*
+ * CONSTANTS, XENSTORE FIELD AND PATH NAME STRINGS, HELPERS.
*/
-/* Pointer movement event */
-#define XENKBD_TYPE_MOTION 1
-/* Event type 2 currently not used */
-/* Key event (includes pointer buttons) */
-#define XENKBD_TYPE_KEY 3
+#define XENKBD_DRIVER_NAME "vkbd"
+
+#define XENKBD_FIELD_FEAT_ABS_POINTER "feature-abs-pointer"
+#define XENKBD_FIELD_REQ_ABS_POINTER "request-abs-pointer"
+#define XENKBD_FIELD_RING_GREF "page-gref"
+#define XENKBD_FIELD_EVT_CHANNEL "event-channel"
+#define XENKBD_FIELD_WIDTH "width"
+#define XENKBD_FIELD_HEIGHT "height"
+
+/* OBSOLETE, not recommended for use */
+#define XENKBD_FIELD_RING_REF "page-ref"
+
/*
- * Pointer position event
- * Capable backend sets feature-abs-pointer in xenstore.
- * Frontend requests ot instead of XENKBD_TYPE_MOTION by setting
- * request-abs-update in xenstore.
+ *****************************************************************************
+ * Description of the protocol between frontend and backend driver.
+ *****************************************************************************
+ *
+ * The two halves of a Para-virtual driver communicate with
+ * each other using a shared page and an event channel.
+ * Shared page contains a ring with event structures.
+ *
+ * All reserved fields in the structures below must be 0.
+ *
+ *****************************************************************************
+ * Backend to frontend events
+ *****************************************************************************
+ *
+ * Frontends should ignore unknown in events.
+ * All event packets have the same length (40 octets)
+ * All event packets have common header:
+ *
+ * 0 octet
+ * +-----------------+
+ * | type |
+ * +-----------------+
+ * type - uint8_t, event code, XENKBD_TYPE_???
+ *
+ *
+ * Pointer relative movement event
+ * 0 1 2 3 octet
+ * +----------------+----------------+----------------+----------------+
+ * | _TYPE_MOTION | reserved | 4
+ * +----------------+----------------+----------------+----------------+
+ * | rel_x | 8
+ * +----------------+----------------+----------------+----------------+
+ * | rel_y | 12
+ * +----------------+----------------+----------------+----------------+
+ * | rel_z | 16
+ * +----------------+----------------+----------------+----------------+
+ * | reserved | 20
+ * +----------------+----------------+----------------+----------------+
+ * |/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/|
+ * +----------------+----------------+----------------+----------------+
+ * | reserved | 40
+ * +----------------+----------------+----------------+----------------+
+ *
+ * rel_x - int32_t, relative X motion
+ * rel_y - int32_t, relative Y motion
+ * rel_z - int32_t, relative Z motion (wheel)
*/
-#define XENKBD_TYPE_POS 4
struct xenkbd_motion {
- uint8_t type; /* XENKBD_TYPE_MOTION */
- int32_t rel_x; /* relative X motion */
- int32_t rel_y; /* relative Y motion */
- int32_t rel_z; /* relative Z motion (wheel) */
+ uint8_t type;
+ int32_t rel_x;
+ int32_t rel_y;
+ int32_t rel_z;
};
+/*
+ * Key event (includes pointer buttons)
+ * 0 1 2 3 octet
+ * +----------------+----------------+----------------+----------------+
+ * | _TYPE_KEY | pressed | reserved | 4
+ * +----------------+----------------+----------------+----------------+
+ * | keycode | 8
+ * +----------------+----------------+----------------+----------------+
+ * | reserved | 12
+ * +----------------+----------------+----------------+----------------+
+ * |/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/|
+ * +----------------+----------------+----------------+----------------+
+ * | reserved | 40
+ * +----------------+----------------+----------------+----------------+
+ *
+ * pressed - uint8_t, 1 if pressed; 0 otherwise
+ * keycode - uint32_t, KEY_* from linux/input.h
+ */
+
struct xenkbd_key {
- uint8_t type; /* XENKBD_TYPE_KEY */
- uint8_t pressed; /* 1 if pressed; 0 otherwise */
- uint32_t keycode; /* KEY_* from linux/input.h */
+ uint8_t type;
+ uint8_t pressed;
+ uint32_t keycode;
};
+/*
+ * Pointer absolute position event
+ * 0 1 2 3 octet
+ * +----------------+----------------+----------------+----------------+
+ * | _TYPE_POS | reserved | 4
+ * +----------------+----------------+----------------+----------------+
+ * | abs_x | 8
+ * +----------------+----------------+----------------+----------------+
+ * | abs_y | 12
+ * +----------------+----------------+----------------+----------------+
+ * | rel_z | 16
+ * +----------------+----------------+----------------+----------------+
+ * | reserved | 20
+ * +----------------+----------------+----------------+----------------+
+ * |/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/|
+ * +----------------+----------------+----------------+----------------+
+ * | reserved | 40
+ * +----------------+----------------+----------------+----------------+
+ *
+ * abs_x - int32_t, absolute X position (in FB pixels)
+ * abs_y - int32_t, absolute Y position (in FB pixels)
+ * rel_z - int32_t, relative Z motion (wheel)
+ */
+
struct xenkbd_position {
- uint8_t type; /* XENKBD_TYPE_POS */
- int32_t abs_x; /* absolute X position (in FB pixels) */
- int32_t abs_y; /* absolute Y position (in FB pixels) */
- int32_t rel_z; /* relative Z motion (wheel) */
+ uint8_t type;
+ int32_t abs_x;
+ int32_t abs_y;
+ int32_t rel_z;
};
#define XENKBD_IN_EVENT_SIZE 40
@@ -75,12 +255,22 @@ union xenkbd_in_event {
char pad[XENKBD_IN_EVENT_SIZE];
};
-/* Out events (frontend -> backend) */
-
/*
+ *****************************************************************************
+ * Frontend to backend events
+ *****************************************************************************
+ *
* Out events may be sent only when requested by backend, and receipt
* of an unknown out event is an error.
* No out events currently defined.
+
+ * All event packets have the same length (40 octets)
+ * All event packets have common header:
+ * 0 octet
+ * +-----------------+
+ * | type |
+ * +-----------------+
+ * type - uint8_t, event code
*/
#define XENKBD_OUT_EVENT_SIZE 40
@@ -90,7 +280,11 @@ union xenkbd_out_event {
char pad[XENKBD_OUT_EVENT_SIZE];
};
-/* shared page */
+/*
+ *****************************************************************************
+ * Shared page
+ *****************************************************************************
+ */
#define XENKBD_IN_RING_SIZE 2048
#define XENKBD_IN_RING_LEN (XENKBD_IN_RING_SIZE / XENKBD_IN_EVENT_SIZE)
@@ -113,4 +307,14 @@ struct xenkbd_page {
uint32_t out_cons, out_prod;
};
-#endif
+#endif /* __XEN_PUBLIC_IO_KBDIF_H__ */
+
+/*
+ * Local variables:
+ * mode: C
+ * c-file-style: "BSD"
+ * c-basic-offset: 4
+ * tab-width: 4
+ * indent-tabs-mode: nil
+ * End:
+ */
--
2.7.4
_______________________________________________
Xen-devel mailing list
Xen-devel@lists.xen.org
https://lists.xen.org/xen-devel
^ permalink raw reply related [flat|nested] 41+ messages in thread
* [For Linux 2/4] xen/kbdif: add multi-touch support
2017-04-07 8:30 [For Linux 0/4] PV protocol headers for Linux Kernel Oleksandr Andrushchenko
` (2 preceding siblings ...)
2017-04-07 8:30 ` [For Linux 2/4] xen/kbdif: add multi-touch support Oleksandr Andrushchenko
@ 2017-04-07 8:30 ` Oleksandr Andrushchenko
2017-04-07 11:30 ` Juergen Gross
2017-04-07 11:30 ` [Xen-devel] " Juergen Gross
2017-04-07 8:30 ` [For Linux 3/4] xen/sndif: add sound-device ABI Oleksandr Andrushchenko
` (5 subsequent siblings)
9 siblings, 2 replies; 41+ messages in thread
From: Oleksandr Andrushchenko @ 2017-04-07 8:30 UTC (permalink / raw)
To: xen-devel
Cc: joculator, al1img, vlad.babchuk, andrii.anisov, olekstysh,
konrad.wilk, lars.kurth, julien.grall, andr2000, sstabellini,
linux-kernel, Oleksandr Andrushchenko
From: Oleksandr Andrushchenko <oleksandr_andrushchenko@epam.com>
Multi-touch fields re-use the page that is used by the other features
which means that you can interleave multi-touch, motion, and key
events.
Reviewed-by: Konrad Rzeszutek Wilk <konrad.wilk@oracle.com>
Reviewed-by: Stefano Stabellini <sstabellini@kernel.org>
Signed-off-by: Oleksandr Andrushchenko <oleksandr_andrushchenko@epam.com>
---
include/xen/interface/io/kbdif.h | 210 +++++++++++++++++++++++++++++++++++++++
1 file changed, 210 insertions(+)
diff --git a/include/xen/interface/io/kbdif.h b/include/xen/interface/io/kbdif.h
index e83eaddb20dc..41124f69aa59 100644
--- a/include/xen/interface/io/kbdif.h
+++ b/include/xen/interface/io/kbdif.h
@@ -57,6 +57,12 @@
* Backends, which support reporting of absolute coordinates for pointer
* device should set this to 1.
*
+ * feature-multi-touch
+ * Values: <uint>
+ *
+ * Backends, which support reporting of multi-touch events
+ * should set this to 1.
+ *
*------------------------- Pointer Device Parameters ------------------------
*
* width
@@ -87,6 +93,11 @@
* Request backend to report absolute pointer coordinates
* (XENKBD_TYPE_POS) instead of relative ones (XENKBD_TYPE_MOTION).
*
+ * request-multi-touch
+ * Values: <uint>
+ *
+ * Request backend to report multi-touch events.
+ *
*----------------------- Request Transport Parameters -----------------------
*
* event-channel
@@ -106,6 +117,25 @@
*
* OBSOLETE, not recommended for use.
* PFN of the shared page.
+ *
+ *----------------------- Multi-touch Device Parameters -----------------------
+ *
+ * multi-touch-num-contacts
+ * Values: <uint>
+ *
+ * Number of simultaneous touches reported.
+ *
+ * multi-touch-width
+ * Values: <uint>
+ *
+ * Width of the touch area to be used by the frontend
+ * while reporting input events, pixels, [0; UINT32_MAX].
+ *
+ * multi-touch-height
+ * Values: <uint>
+ *
+ * Height of the touch area to be used by the frontend
+ * while reporting input events, pixels, [0; UINT32_MAX].
*/
/*
@@ -116,6 +146,16 @@
#define XENKBD_TYPE_RESERVED 2
#define XENKBD_TYPE_KEY 3
#define XENKBD_TYPE_POS 4
+#define XENKBD_TYPE_MTOUCH 5
+
+/* Multi-touch event sub-codes */
+
+#define XENKBD_MT_EV_DOWN 0
+#define XENKBD_MT_EV_UP 1
+#define XENKBD_MT_EV_MOTION 2
+#define XENKBD_MT_EV_SYN 3
+#define XENKBD_MT_EV_SHAPE 4
+#define XENKBD_MT_EV_ORIENT 5
/*
* CONSTANTS, XENSTORE FIELD AND PATH NAME STRINGS, HELPERS.
@@ -124,11 +164,16 @@
#define XENKBD_DRIVER_NAME "vkbd"
#define XENKBD_FIELD_FEAT_ABS_POINTER "feature-abs-pointer"
+#define XENKBD_FIELD_FEAT_MTOUCH "feature-multi-touch"
#define XENKBD_FIELD_REQ_ABS_POINTER "request-abs-pointer"
+#define XENKBD_FIELD_REQ_MTOUCH "request-multi-touch"
#define XENKBD_FIELD_RING_GREF "page-gref"
#define XENKBD_FIELD_EVT_CHANNEL "event-channel"
#define XENKBD_FIELD_WIDTH "width"
#define XENKBD_FIELD_HEIGHT "height"
+#define XENKBD_FIELD_MT_WIDTH "multi-touch-width"
+#define XENKBD_FIELD_MT_HEIGHT "multi-touch-height"
+#define XENKBD_FIELD_MT_NUM_CONTACTS "multi-touch-num-contacts"
/* OBSOLETE, not recommended for use */
#define XENKBD_FIELD_RING_REF "page-ref"
@@ -245,6 +290,170 @@ struct xenkbd_position {
int32_t rel_z;
};
+/*
+ * Multi-touch event and its sub-types
+ *
+ * All multi-touch event packets have common header:
+ *
+ * 0 1 2 3 octet
+ * +----------------+----------------+----------------+----------------+
+ * | _TYPE_MTOUCH | event_type | contact_id | reserved | 4
+ * +----------------+----------------+----------------+----------------+
+ * | reserved | 8
+ * +----------------+----------------+----------------+----------------+
+ *
+ * event_type - unt8_t, multi-touch event sub-type, XENKBD_MT_EV_???
+ * contact_id - unt8_t, ID of the contact
+ *
+ * Touch interactions can consist of one or more contacts.
+ * For each contact, a series of events is generated, starting
+ * with a down event, followed by zero or more motion events,
+ * and ending with an up event. Events relating to the same
+ * contact point can be identified by the ID of the sequence: contact ID.
+ * Contact ID may be reused after XENKBD_MT_EV_UP event and
+ * is in the [0; XENKBD_FIELD_NUM_CONTACTS - 1] range.
+ *
+ * For further information please refer to documentation on Wayland [1],
+ * Linux [2] and Windows [3] multi-touch support.
+ *
+ * [1] https://cgit.freedesktop.org/wayland/wayland/tree/protocol/wayland.xml
+ * [2] https://www.kernel.org/doc/Documentation/input/multi-touch-protocol.txt
+ * [3] https://msdn.microsoft.com/en-us/library/jj151564(v=vs.85).aspx
+ *
+ *
+ * Multi-touch down event - sent when a new touch is made: touch is assigned
+ * a unique contact ID, sent with this and consequent events related
+ * to this touch.
+ * 0 1 2 3 octet
+ * +----------------+----------------+----------------+----------------+
+ * | _TYPE_MTOUCH | _MT_EV_DOWN | contact_id | reserved | 4
+ * +----------------+----------------+----------------+----------------+
+ * | reserved | 8
+ * +----------------+----------------+----------------+----------------+
+ * | abs_x | 12
+ * +----------------+----------------+----------------+----------------+
+ * | abs_y | 16
+ * +----------------+----------------+----------------+----------------+
+ * | reserved | 20
+ * +----------------+----------------+----------------+----------------+
+ * |/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/|
+ * +----------------+----------------+----------------+----------------+
+ * | reserved | 40
+ * +----------------+----------------+----------------+----------------+
+ *
+ * abs_x - int32_t, absolute X position, in pixels
+ * abs_y - int32_t, absolute Y position, in pixels
+ *
+ * Multi-touch contact release event
+ * 0 1 2 3 octet
+ * +----------------+----------------+----------------+----------------+
+ * | _TYPE_MTOUCH | _MT_EV_UP | contact_id | reserved | 4
+ * +----------------+----------------+----------------+----------------+
+ * | reserved | 8
+ * +----------------+----------------+----------------+----------------+
+ * |/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/|
+ * +----------------+----------------+----------------+----------------+
+ * | reserved | 40
+ * +----------------+----------------+----------------+----------------+
+ *
+ * Multi-touch motion event
+ * 0 1 2 3 octet
+ * +----------------+----------------+----------------+----------------+
+ * | _TYPE_MTOUCH | _MT_EV_MOTION | contact_id | reserved | 4
+ * +----------------+----------------+----------------+----------------+
+ * | reserved | 8
+ * +----------------+----------------+----------------+----------------+
+ * | abs_x | 12
+ * +----------------+----------------+----------------+----------------+
+ * | abs_y | 16
+ * +----------------+----------------+----------------+----------------+
+ * | reserved | 20
+ * +----------------+----------------+----------------+----------------+
+ * |/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/|
+ * +----------------+----------------+----------------+----------------+
+ * | reserved | 40
+ * +----------------+----------------+----------------+----------------+
+ *
+ * abs_x - int32_t, absolute X position, in pixels,
+ * abs_y - int32_t, absolute Y position, in pixels,
+ *
+ * Multi-touch input synchronization event - shows end of a set of events
+ * which logically belong together.
+ * 0 1 2 3 octet
+ * +----------------+----------------+----------------+----------------+
+ * | _TYPE_MTOUCH | _MT_EV_SYN | contact_id | reserved | 4
+ * +----------------+----------------+----------------+----------------+
+ * | reserved | 8
+ * +----------------+----------------+----------------+----------------+
+ * |/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/|
+ * +----------------+----------------+----------------+----------------+
+ * | reserved | 40
+ * +----------------+----------------+----------------+----------------+
+ *
+ * Multi-touch shape event - touch point's shape has changed its shape.
+ * Shape is approximated by an ellipse through the major and minor axis
+ * lengths: major is the longer diameter of the ellipse and minor is the
+ * shorter one. Center of the ellipse is reported via
+ * XENKBD_MT_EV_DOWN/XENKBD_MT_EV_MOTION events.
+ * 0 1 2 3 octet
+ * +----------------+----------------+----------------+----------------+
+ * | _TYPE_MTOUCH | _MT_EV_SHAPE | contact_id | reserved | 4
+ * +----------------+----------------+----------------+----------------+
+ * | reserved | 8
+ * +----------------+----------------+----------------+----------------+
+ * | major | 12
+ * +----------------+----------------+----------------+----------------+
+ * | minor | 16
+ * +----------------+----------------+----------------+----------------+
+ * | reserved | 20
+ * +----------------+----------------+----------------+----------------+
+ * |/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/|
+ * +----------------+----------------+----------------+----------------+
+ * | reserved | 40
+ * +----------------+----------------+----------------+----------------+
+ *
+ * major - unt32_t, length of the major axis, pixels
+ * minor - unt32_t, length of the minor axis, pixels
+ *
+ * Multi-touch orientation event - touch point's shape has changed
+ * its orientation: calculated as a clockwise angle between the major axis
+ * of the ellipse and positive Y axis in degrees, [-180; +180].
+ * 0 1 2 3 octet
+ * +----------------+----------------+----------------+----------------+
+ * | _TYPE_MTOUCH | _MT_EV_ORIENT | contact_id | reserved | 4
+ * +----------------+----------------+----------------+----------------+
+ * | reserved | 8
+ * +----------------+----------------+----------------+----------------+
+ * | orientation | reserved | 12
+ * +----------------+----------------+----------------+----------------+
+ * | reserved | 16
+ * +----------------+----------------+----------------+----------------+
+ * |/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/|
+ * +----------------+----------------+----------------+----------------+
+ * | reserved | 40
+ * +----------------+----------------+----------------+----------------+
+ *
+ * orientation - int16_t, clockwise angle of the major axis
+ */
+
+struct xenkbd_mtouch {
+ uint8_t type; /* XENKBD_TYPE_MTOUCH */
+ uint8_t event_type; /* XENKBD_MT_EV_??? */
+ uint8_t contact_id;
+ uint8_t reserved[5]; /* reserved for the future use */
+ union {
+ struct {
+ int32_t abs_x; /* absolute X position, pixels */
+ int32_t abs_y; /* absolute Y position, pixels */
+ } pos;
+ struct {
+ uint32_t major; /* length of the major axis, pixels */
+ uint32_t minor; /* length of the minor axis, pixels */
+ } shape;
+ int16_t orientation; /* clockwise angle of the major axis */
+ } u;
+};
+
#define XENKBD_IN_EVENT_SIZE 40
union xenkbd_in_event {
@@ -252,6 +461,7 @@ union xenkbd_in_event {
struct xenkbd_motion motion;
struct xenkbd_key key;
struct xenkbd_position pos;
+ struct xenkbd_mtouch mtouch;
char pad[XENKBD_IN_EVENT_SIZE];
};
--
2.7.4
^ permalink raw reply related [flat|nested] 41+ messages in thread
* [For Linux 2/4] xen/kbdif: add multi-touch support
2017-04-07 8:30 [For Linux 0/4] PV protocol headers for Linux Kernel Oleksandr Andrushchenko
2017-04-07 8:30 ` [For Linux 1/4] xen/kbdif: update protocol description Oleksandr Andrushchenko
2017-04-07 8:30 ` Oleksandr Andrushchenko
@ 2017-04-07 8:30 ` Oleksandr Andrushchenko
2017-04-07 8:30 ` Oleksandr Andrushchenko
` (6 subsequent siblings)
9 siblings, 0 replies; 41+ messages in thread
From: Oleksandr Andrushchenko @ 2017-04-07 8:30 UTC (permalink / raw)
To: xen-devel
Cc: lars.kurth, sstabellini, vlad.babchuk, linux-kernel,
Oleksandr Andrushchenko, julien.grall, andrii.anisov, olekstysh,
andr2000, al1img, joculator
From: Oleksandr Andrushchenko <oleksandr_andrushchenko@epam.com>
Multi-touch fields re-use the page that is used by the other features
which means that you can interleave multi-touch, motion, and key
events.
Reviewed-by: Konrad Rzeszutek Wilk <konrad.wilk@oracle.com>
Reviewed-by: Stefano Stabellini <sstabellini@kernel.org>
Signed-off-by: Oleksandr Andrushchenko <oleksandr_andrushchenko@epam.com>
---
include/xen/interface/io/kbdif.h | 210 +++++++++++++++++++++++++++++++++++++++
1 file changed, 210 insertions(+)
diff --git a/include/xen/interface/io/kbdif.h b/include/xen/interface/io/kbdif.h
index e83eaddb20dc..41124f69aa59 100644
--- a/include/xen/interface/io/kbdif.h
+++ b/include/xen/interface/io/kbdif.h
@@ -57,6 +57,12 @@
* Backends, which support reporting of absolute coordinates for pointer
* device should set this to 1.
*
+ * feature-multi-touch
+ * Values: <uint>
+ *
+ * Backends, which support reporting of multi-touch events
+ * should set this to 1.
+ *
*------------------------- Pointer Device Parameters ------------------------
*
* width
@@ -87,6 +93,11 @@
* Request backend to report absolute pointer coordinates
* (XENKBD_TYPE_POS) instead of relative ones (XENKBD_TYPE_MOTION).
*
+ * request-multi-touch
+ * Values: <uint>
+ *
+ * Request backend to report multi-touch events.
+ *
*----------------------- Request Transport Parameters -----------------------
*
* event-channel
@@ -106,6 +117,25 @@
*
* OBSOLETE, not recommended for use.
* PFN of the shared page.
+ *
+ *----------------------- Multi-touch Device Parameters -----------------------
+ *
+ * multi-touch-num-contacts
+ * Values: <uint>
+ *
+ * Number of simultaneous touches reported.
+ *
+ * multi-touch-width
+ * Values: <uint>
+ *
+ * Width of the touch area to be used by the frontend
+ * while reporting input events, pixels, [0; UINT32_MAX].
+ *
+ * multi-touch-height
+ * Values: <uint>
+ *
+ * Height of the touch area to be used by the frontend
+ * while reporting input events, pixels, [0; UINT32_MAX].
*/
/*
@@ -116,6 +146,16 @@
#define XENKBD_TYPE_RESERVED 2
#define XENKBD_TYPE_KEY 3
#define XENKBD_TYPE_POS 4
+#define XENKBD_TYPE_MTOUCH 5
+
+/* Multi-touch event sub-codes */
+
+#define XENKBD_MT_EV_DOWN 0
+#define XENKBD_MT_EV_UP 1
+#define XENKBD_MT_EV_MOTION 2
+#define XENKBD_MT_EV_SYN 3
+#define XENKBD_MT_EV_SHAPE 4
+#define XENKBD_MT_EV_ORIENT 5
/*
* CONSTANTS, XENSTORE FIELD AND PATH NAME STRINGS, HELPERS.
@@ -124,11 +164,16 @@
#define XENKBD_DRIVER_NAME "vkbd"
#define XENKBD_FIELD_FEAT_ABS_POINTER "feature-abs-pointer"
+#define XENKBD_FIELD_FEAT_MTOUCH "feature-multi-touch"
#define XENKBD_FIELD_REQ_ABS_POINTER "request-abs-pointer"
+#define XENKBD_FIELD_REQ_MTOUCH "request-multi-touch"
#define XENKBD_FIELD_RING_GREF "page-gref"
#define XENKBD_FIELD_EVT_CHANNEL "event-channel"
#define XENKBD_FIELD_WIDTH "width"
#define XENKBD_FIELD_HEIGHT "height"
+#define XENKBD_FIELD_MT_WIDTH "multi-touch-width"
+#define XENKBD_FIELD_MT_HEIGHT "multi-touch-height"
+#define XENKBD_FIELD_MT_NUM_CONTACTS "multi-touch-num-contacts"
/* OBSOLETE, not recommended for use */
#define XENKBD_FIELD_RING_REF "page-ref"
@@ -245,6 +290,170 @@ struct xenkbd_position {
int32_t rel_z;
};
+/*
+ * Multi-touch event and its sub-types
+ *
+ * All multi-touch event packets have common header:
+ *
+ * 0 1 2 3 octet
+ * +----------------+----------------+----------------+----------------+
+ * | _TYPE_MTOUCH | event_type | contact_id | reserved | 4
+ * +----------------+----------------+----------------+----------------+
+ * | reserved | 8
+ * +----------------+----------------+----------------+----------------+
+ *
+ * event_type - unt8_t, multi-touch event sub-type, XENKBD_MT_EV_???
+ * contact_id - unt8_t, ID of the contact
+ *
+ * Touch interactions can consist of one or more contacts.
+ * For each contact, a series of events is generated, starting
+ * with a down event, followed by zero or more motion events,
+ * and ending with an up event. Events relating to the same
+ * contact point can be identified by the ID of the sequence: contact ID.
+ * Contact ID may be reused after XENKBD_MT_EV_UP event and
+ * is in the [0; XENKBD_FIELD_NUM_CONTACTS - 1] range.
+ *
+ * For further information please refer to documentation on Wayland [1],
+ * Linux [2] and Windows [3] multi-touch support.
+ *
+ * [1] https://cgit.freedesktop.org/wayland/wayland/tree/protocol/wayland.xml
+ * [2] https://www.kernel.org/doc/Documentation/input/multi-touch-protocol.txt
+ * [3] https://msdn.microsoft.com/en-us/library/jj151564(v=vs.85).aspx
+ *
+ *
+ * Multi-touch down event - sent when a new touch is made: touch is assigned
+ * a unique contact ID, sent with this and consequent events related
+ * to this touch.
+ * 0 1 2 3 octet
+ * +----------------+----------------+----------------+----------------+
+ * | _TYPE_MTOUCH | _MT_EV_DOWN | contact_id | reserved | 4
+ * +----------------+----------------+----------------+----------------+
+ * | reserved | 8
+ * +----------------+----------------+----------------+----------------+
+ * | abs_x | 12
+ * +----------------+----------------+----------------+----------------+
+ * | abs_y | 16
+ * +----------------+----------------+----------------+----------------+
+ * | reserved | 20
+ * +----------------+----------------+----------------+----------------+
+ * |/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/|
+ * +----------------+----------------+----------------+----------------+
+ * | reserved | 40
+ * +----------------+----------------+----------------+----------------+
+ *
+ * abs_x - int32_t, absolute X position, in pixels
+ * abs_y - int32_t, absolute Y position, in pixels
+ *
+ * Multi-touch contact release event
+ * 0 1 2 3 octet
+ * +----------------+----------------+----------------+----------------+
+ * | _TYPE_MTOUCH | _MT_EV_UP | contact_id | reserved | 4
+ * +----------------+----------------+----------------+----------------+
+ * | reserved | 8
+ * +----------------+----------------+----------------+----------------+
+ * |/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/|
+ * +----------------+----------------+----------------+----------------+
+ * | reserved | 40
+ * +----------------+----------------+----------------+----------------+
+ *
+ * Multi-touch motion event
+ * 0 1 2 3 octet
+ * +----------------+----------------+----------------+----------------+
+ * | _TYPE_MTOUCH | _MT_EV_MOTION | contact_id | reserved | 4
+ * +----------------+----------------+----------------+----------------+
+ * | reserved | 8
+ * +----------------+----------------+----------------+----------------+
+ * | abs_x | 12
+ * +----------------+----------------+----------------+----------------+
+ * | abs_y | 16
+ * +----------------+----------------+----------------+----------------+
+ * | reserved | 20
+ * +----------------+----------------+----------------+----------------+
+ * |/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/|
+ * +----------------+----------------+----------------+----------------+
+ * | reserved | 40
+ * +----------------+----------------+----------------+----------------+
+ *
+ * abs_x - int32_t, absolute X position, in pixels,
+ * abs_y - int32_t, absolute Y position, in pixels,
+ *
+ * Multi-touch input synchronization event - shows end of a set of events
+ * which logically belong together.
+ * 0 1 2 3 octet
+ * +----------------+----------------+----------------+----------------+
+ * | _TYPE_MTOUCH | _MT_EV_SYN | contact_id | reserved | 4
+ * +----------------+----------------+----------------+----------------+
+ * | reserved | 8
+ * +----------------+----------------+----------------+----------------+
+ * |/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/|
+ * +----------------+----------------+----------------+----------------+
+ * | reserved | 40
+ * +----------------+----------------+----------------+----------------+
+ *
+ * Multi-touch shape event - touch point's shape has changed its shape.
+ * Shape is approximated by an ellipse through the major and minor axis
+ * lengths: major is the longer diameter of the ellipse and minor is the
+ * shorter one. Center of the ellipse is reported via
+ * XENKBD_MT_EV_DOWN/XENKBD_MT_EV_MOTION events.
+ * 0 1 2 3 octet
+ * +----------------+----------------+----------------+----------------+
+ * | _TYPE_MTOUCH | _MT_EV_SHAPE | contact_id | reserved | 4
+ * +----------------+----------------+----------------+----------------+
+ * | reserved | 8
+ * +----------------+----------------+----------------+----------------+
+ * | major | 12
+ * +----------------+----------------+----------------+----------------+
+ * | minor | 16
+ * +----------------+----------------+----------------+----------------+
+ * | reserved | 20
+ * +----------------+----------------+----------------+----------------+
+ * |/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/|
+ * +----------------+----------------+----------------+----------------+
+ * | reserved | 40
+ * +----------------+----------------+----------------+----------------+
+ *
+ * major - unt32_t, length of the major axis, pixels
+ * minor - unt32_t, length of the minor axis, pixels
+ *
+ * Multi-touch orientation event - touch point's shape has changed
+ * its orientation: calculated as a clockwise angle between the major axis
+ * of the ellipse and positive Y axis in degrees, [-180; +180].
+ * 0 1 2 3 octet
+ * +----------------+----------------+----------------+----------------+
+ * | _TYPE_MTOUCH | _MT_EV_ORIENT | contact_id | reserved | 4
+ * +----------------+----------------+----------------+----------------+
+ * | reserved | 8
+ * +----------------+----------------+----------------+----------------+
+ * | orientation | reserved | 12
+ * +----------------+----------------+----------------+----------------+
+ * | reserved | 16
+ * +----------------+----------------+----------------+----------------+
+ * |/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/|
+ * +----------------+----------------+----------------+----------------+
+ * | reserved | 40
+ * +----------------+----------------+----------------+----------------+
+ *
+ * orientation - int16_t, clockwise angle of the major axis
+ */
+
+struct xenkbd_mtouch {
+ uint8_t type; /* XENKBD_TYPE_MTOUCH */
+ uint8_t event_type; /* XENKBD_MT_EV_??? */
+ uint8_t contact_id;
+ uint8_t reserved[5]; /* reserved for the future use */
+ union {
+ struct {
+ int32_t abs_x; /* absolute X position, pixels */
+ int32_t abs_y; /* absolute Y position, pixels */
+ } pos;
+ struct {
+ uint32_t major; /* length of the major axis, pixels */
+ uint32_t minor; /* length of the minor axis, pixels */
+ } shape;
+ int16_t orientation; /* clockwise angle of the major axis */
+ } u;
+};
+
#define XENKBD_IN_EVENT_SIZE 40
union xenkbd_in_event {
@@ -252,6 +461,7 @@ union xenkbd_in_event {
struct xenkbd_motion motion;
struct xenkbd_key key;
struct xenkbd_position pos;
+ struct xenkbd_mtouch mtouch;
char pad[XENKBD_IN_EVENT_SIZE];
};
--
2.7.4
_______________________________________________
Xen-devel mailing list
Xen-devel@lists.xen.org
https://lists.xen.org/xen-devel
^ permalink raw reply related [flat|nested] 41+ messages in thread
* [For Linux 3/4] xen/sndif: add sound-device ABI
2017-04-07 8:30 [For Linux 0/4] PV protocol headers for Linux Kernel Oleksandr Andrushchenko
` (4 preceding siblings ...)
2017-04-07 8:30 ` [For Linux 3/4] xen/sndif: add sound-device ABI Oleksandr Andrushchenko
@ 2017-04-07 8:30 ` Oleksandr Andrushchenko
2017-04-07 11:31 ` Juergen Gross
2017-04-07 11:31 ` [Xen-devel] " Juergen Gross
2017-04-07 8:30 ` [For Linux 4/4] xen/displif: add ABI for para-virtual display Oleksandr Andrushchenko
` (3 subsequent siblings)
9 siblings, 2 replies; 41+ messages in thread
From: Oleksandr Andrushchenko @ 2017-04-07 8:30 UTC (permalink / raw)
To: xen-devel
Cc: joculator, al1img, vlad.babchuk, andrii.anisov, olekstysh,
konrad.wilk, lars.kurth, julien.grall, andr2000, sstabellini,
linux-kernel, Oleksandr Andrushchenko, Oleksandr Grytsov,
Oleksandr Dmytryshyn, Iurii Konovalenko
From: Oleksandr Andrushchenko <oleksandr_andrushchenko@epam.com>
Add ABI for the two halves of a para-virtualized
sound driver to communicate with each other.
The ABI allows implementing audio playback and capture as
well as volume control and possibility to mute/unmute
audio sources.
Note: depending on the use-case backend can expose more sound
cards and PCM devices/streams than the underlying HW physically
has by employing SW mixers, configuring virtual sound streams,
channels etc. Thus, allowing fine tunned configurations per
frontend.
Reviewed-by: Konrad Rzeszutek Wilk <konrad.wilk@oracle.com>
Signed-off-by: Oleksandr Andrushchenko <oleksandr_andrushchenko@epam.com>
Signed-off-by: Oleksandr Grytsov <oleksandr_grytsov@epam.com>
Signed-off-by: Oleksandr Dmytryshyn <oleksandr.dmytryshyn@globallogic.com>
Signed-off-by: Iurii Konovalenko <iurii.konovalenko@globallogic.com>
---
include/xen/interface/io/sndif.h | 803 +++++++++++++++++++++++++++++++++++++++
1 file changed, 803 insertions(+)
create mode 100644 include/xen/interface/io/sndif.h
diff --git a/include/xen/interface/io/sndif.h b/include/xen/interface/io/sndif.h
new file mode 100644
index 000000000000..ea9b81024cd9
--- /dev/null
+++ b/include/xen/interface/io/sndif.h
@@ -0,0 +1,803 @@
+/******************************************************************************
+ * sndif.h
+ *
+ * Unified sound-device I/O interface for Xen guest OSes.
+ *
+ * Permission is hereby granted, free of charge, to any person obtaining a copy
+ * of this software and associated documentation files (the "Software"), to
+ * deal in the Software without restriction, including without limitation the
+ * rights to use, copy, modify, merge, publish, distribute, sublicense, and/or
+ * sell copies of the Software, and to permit persons to whom the Software is
+ * furnished to do so, subject to the following conditions:
+ *
+ * The above copyright notice and this permission notice shall be included in
+ * all copies or substantial portions of the Software.
+ *
+ * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
+ * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
+ * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE
+ * AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER
+ * LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING
+ * FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER
+ * DEALINGS IN THE SOFTWARE.
+ *
+ * Copyright (C) 2013-2015 GlobalLogic Inc.
+ * Copyright (C) 2016-2017 EPAM Systems Inc.
+ *
+ * Authors: Oleksandr Andrushchenko <oleksandr_andrushchenko@epam.com>
+ * Oleksandr Grytsov <oleksandr_grytsov@epam.com>
+ * Oleksandr Dmytryshyn <oleksandr.dmytryshyn@globallogic.com>
+ * Iurii Konovalenko <iurii.konovalenko@globallogic.com>
+ */
+
+#ifndef __XEN_PUBLIC_IO_SNDIF_H__
+#define __XEN_PUBLIC_IO_SNDIF_H__
+
+#include "ring.h"
+#include "../grant_table.h"
+
+/*
+ ******************************************************************************
+ * Feature and Parameter Negotiation
+ ******************************************************************************
+ *
+ * Front->back notifications: when enqueuing a new request, sending a
+ * notification can be made conditional on xensnd_req (i.e., the generic
+ * hold-off mechanism provided by the ring macros). Backends must set
+ * xensnd_req appropriately (e.g., using RING_FINAL_CHECK_FOR_REQUESTS()).
+ *
+ * Back->front notifications: when enqueuing a new response, sending a
+ * notification can be made conditional on xensnd_resp (i.e., the generic
+ * hold-off mechanism provided by the ring macros). Frontends must set
+ * xensnd_resp appropriately (e.g., using RING_FINAL_CHECK_FOR_RESPONSES()).
+ *
+ * The two halves of a para-virtual sound card driver utilize nodes within
+ * XenStore to communicate capabilities and to negotiate operating parameters.
+ * This section enumerates these nodes which reside in the respective front and
+ * backend portions of XenStore, following the XenBus convention.
+ *
+ * All data in XenStore is stored as strings. Nodes specifying numeric
+ * values are encoded in decimal. Integer value ranges listed below are
+ * expressed as fixed sized integer types capable of storing the conversion
+ * of a properly formated node string, without loss of information.
+ *
+ ******************************************************************************
+ * Example configuration
+ ******************************************************************************
+ *
+ * Note: depending on the use-case backend can expose more sound cards and
+ * PCM devices/streams than the underlying HW physically has by employing
+ * SW mixers, configuring virtual sound streams, channels etc.
+ *
+ * This is an example of backend and frontend configuration:
+ *
+ *--------------------------------- Backend -----------------------------------
+ *
+ * /local/domain/0/backend/vsnd/1/0/frontend-id = "1"
+ * /local/domain/0/backend/vsnd/1/0/frontend = "/local/domain/1/device/vsnd/0"
+ * /local/domain/0/backend/vsnd/1/0/state = "4"
+ * /local/domain/0/backend/vsnd/1/0/versions = "1,2"
+ *
+ *--------------------------------- Frontend ----------------------------------
+ *
+ * /local/domain/1/device/vsnd/0/backend-id = "0"
+ * /local/domain/1/device/vsnd/0/backend = "/local/domain/0/backend/vsnd/1/0"
+ * /local/domain/1/device/vsnd/0/state = "4"
+ * /local/domain/1/device/vsnd/0/version = "1"
+ *
+ *----------------------------- Card configuration ----------------------------
+ *
+ * /local/domain/1/device/vsnd/0/short-name = "Card short name"
+ * /local/domain/1/device/vsnd/0/long-name = "Card long name"
+ * /local/domain/1/device/vsnd/0/sample-rates = "8000,32000,44100,48000,96000"
+ * /local/domain/1/device/vsnd/0/sample-formats = "s8,u8,s16_le,s16_be"
+ * /local/domain/1/device/vsnd/0/buffer-size = "262144"
+ *
+ *------------------------------- PCM device 0 --------------------------------
+ *
+ * /local/domain/1/device/vsnd/0/0/name = "General analog"
+ * /local/domain/1/device/vsnd/0/0/channels-max = "5"
+ *
+ *----------------------------- Stream 0, playback ----------------------------
+ *
+ * /local/domain/1/device/vsnd/0/0/0/type = "p"
+ * /local/domain/1/device/vsnd/0/0/0/sample-formats = "s8,u8"
+ * /local/domain/1/device/vsnd/0/0/0/unique-id = "0"
+ *
+ * /local/domain/1/device/vsnd/0/0/0/ring-ref = "386"
+ * /local/domain/1/device/vsnd/0/0/0/event-channel = "15"
+ *
+ *------------------------------ Stream 1, capture ----------------------------
+ *
+ * /local/domain/1/device/vsnd/0/0/1/type = "c"
+ * /local/domain/1/device/vsnd/0/0/1/channels-max = "2"
+ * /local/domain/1/device/vsnd/0/0/1/unique-id = "1"
+ *
+ * /local/domain/1/device/vsnd/0/0/1/ring-ref = "384"
+ * /local/domain/1/device/vsnd/0/0/1/event-channel = "13"
+ *
+ *------------------------------- PCM device 1 --------------------------------
+ *
+ * /local/domain/1/device/vsnd/0/1/name = "HDMI-0"
+ * /local/domain/1/device/vsnd/0/1/sample-rates = "8000,32000,44100"
+ *
+ *------------------------------ Stream 0, capture ----------------------------
+ *
+ * /local/domain/1/device/vsnd/0/1/0/type = "c"
+ * /local/domain/1/device/vsnd/0/1/0/unique-id = "2"
+ *
+ * /local/domain/1/device/vsnd/0/1/0/ring-ref = "387"
+ * /local/domain/1/device/vsnd/0/1/0/event-channel = "151"
+ *
+ *------------------------------- PCM device 2 --------------------------------
+ *
+ * /local/domain/1/device/vsnd/0/2/name = "SPDIF"
+ *
+ *----------------------------- Stream 0, playback ----------------------------
+ *
+ * /local/domain/1/device/vsnd/0/2/0/type = "p"
+ * /local/domain/1/device/vsnd/0/2/0/unique-id = "3"
+ *
+ * /local/domain/1/device/vsnd/0/2/0/ring-ref = "389"
+ * /local/domain/1/device/vsnd/0/2/0/event-channel = "152"
+ *
+ ******************************************************************************
+ * Backend XenBus Nodes
+ ******************************************************************************
+ *
+ *----------------------------- Protocol version ------------------------------
+ *
+ * versions
+ * Values: <string>
+ *
+ * List of XENSND_LIST_SEPARATOR separated protocol versions supported
+ * by the backend. For example "1,2,3".
+ *
+ ******************************************************************************
+ * Frontend XenBus Nodes
+ ******************************************************************************
+ *
+ *-------------------------------- Addressing ---------------------------------
+ *
+ * dom-id
+ * Values: <uint16_t>
+ *
+ * Domain identifier.
+ *
+ * dev-id
+ * Values: <uint16_t>
+ *
+ * Device identifier.
+ *
+ * pcm-dev-idx
+ * Values: <uint8_t>
+ *
+ * Zero based contigous index of the PCM device.
+ *
+ * stream-idx
+ * Values: <uint8_t>
+ *
+ * Zero based contigous index of the stream of the PCM device.
+ *
+ * The following pattern is used for addressing:
+ * /local/domain/<dom-id>/device/vsnd/<dev-id>/<pcm-dev-idx>/<stream-idx>/...
+ *
+ *----------------------------- Protocol version ------------------------------
+ *
+ * version
+ * Values: <string>
+ *
+ * Protocol version, chosen among the ones supported by the backend.
+ *
+ *------------------------------- PCM settings --------------------------------
+ *
+ * Every virtualized sound frontend has a set of PCM devices and streams, each
+ * could be individually configured. Part of the PCM configuration can be
+ * defined at higher level of the hierarchy and be fully or partially re-used
+ * by the underlying layers. These configuration values are:
+ * o number of channels (min/max)
+ * o supported sample rates
+ * o supported sample formats.
+ * E.g. one can define these values for the whole card, device or stream.
+ * Every underlying layer in turn can re-define some or all of them to better
+ * fit its needs. For example, card may define number of channels to be
+ * in [1; 8] range, and some particular stream may be limited to [1; 2] only.
+ * The rule is that the underlying layer must be a subset of the upper layer
+ * range.
+ *
+ * channels-min
+ * Values: <uint8_t>
+ *
+ * The minimum amount of channels that is supported, [1; channels-max].
+ * Optional, if not set or omitted a value of 1 is used.
+ *
+ * channels-max
+ * Values: <uint8_t>
+ *
+ * The maximum amount of channels that is supported.
+ * Must be at least <channels-min>.
+ *
+ * sample-rates
+ * Values: <list of uint32_t>
+ *
+ * List of supported sample rates separated by XENSND_LIST_SEPARATOR.
+ * Sample rates are expressed as a list of decimal values w/o any
+ * ordering requirement.
+ *
+ * sample-formats
+ * Values: <list of XENSND_PCM_FORMAT_XXX_STR>
+ *
+ * List of supported sample formats separated by XENSND_LIST_SEPARATOR.
+ * Items must not exceed XENSND_SAMPLE_FORMAT_MAX_LEN length.
+ *
+ * buffer-size
+ * Values: <uint32_t>
+ *
+ * The maximum size in octets of the buffer to allocate per stream.
+ *
+ *----------------------- Virtual sound card settings -------------------------
+ * short-name
+ * Values: <char[32]>
+ *
+ * Short name of the virtual sound card. Optional.
+ *
+ * long-name
+ * Values: <char[80]>
+ *
+ * Long name of the virtual sound card. Optional.
+ *
+ *----------------------------- Device settings -------------------------------
+ * name
+ * Values: <char[80]>
+ *
+ * Name of the sound device within the virtual sound card. Optional.
+ *
+ *----------------------------- Stream settings -------------------------------
+ *
+ * type
+ * Values: "p", "c"
+ *
+ * Stream type: "p" - playback stream, "c" - capture stream
+ *
+ * If both capture and playback are needed then two streams need to be
+ * defined under the same device.
+ *
+ * unique-id
+ * Values: <uint32_t>
+ *
+ * After stream initialization it is assigned a unique ID (within the front
+ * driver), so every stream of the frontend can be identified by the
+ * backend by this ID. This is not equal to stream-idx as the later is
+ * zero based within the device, but this index is contigous within the
+ * driver.
+ *
+ *-------------------- Stream Request Transport Parameters --------------------
+ *
+ * event-channel
+ * Values: <uint32_t>
+ *
+ * The identifier of the Xen event channel used to signal activity
+ * in the ring buffer.
+ *
+ * ring-ref
+ * Values: <uint32_t>
+ *
+ * The Xen grant reference granting permission for the backend to map
+ * a sole page in a single page sized ring buffer.
+ *
+ ******************************************************************************
+ * STATE DIAGRAMS
+ ******************************************************************************
+ *
+ * Tool stack creates front and back state nodes with initial state
+ * XenbusStateInitialising.
+ * Tool stack creates and sets up frontend sound configuration nodes per domain.
+ *
+ * Front Back
+ * ================================= =====================================
+ * XenbusStateInitialising XenbusStateInitialising
+ * o Query backend device identification
+ * data.
+ * o Open and validate backend device.
+ * |
+ * |
+ * V
+ * XenbusStateInitWait
+ *
+ * o Query frontend configuration
+ * o Allocate and initialize
+ * event channels per configured
+ * playback/capture stream.
+ * o Publish transport parameters
+ * that will be in effect during
+ * this connection.
+ * |
+ * |
+ * V
+ * XenbusStateInitialised
+ *
+ * o Query frontend transport parameters.
+ * o Connect to the event channels.
+ * |
+ * |
+ * V
+ * XenbusStateConnected
+ *
+ * o Create and initialize OS
+ * virtual sound device instances
+ * as per configuration.
+ * |
+ * |
+ * V
+ * XenbusStateConnected
+ *
+ * XenbusStateUnknown
+ * XenbusStateClosed
+ * XenbusStateClosing
+ * o Remove virtual sound device
+ * o Remove event channels
+ * |
+ * |
+ * V
+ * XenbusStateClosed
+ *
+ *------------------------------- Recovery flow -------------------------------
+ *
+ * In case of frontend unrecoverable errors backend handles that as
+ * if frontend goes into the XenbusStateClosed state.
+ *
+ * In case of backend unrecoverable errors frontend tries removing
+ * the virtualized device. If this is possible at the moment of error,
+ * then frontend goes into the XenbusStateInitialising state and is ready for
+ * new connection with backend. If the virtualized device is still in use and
+ * cannot be removed, then frontend goes into the XenbusStateReconfiguring state
+ * until either the virtualized device removed or backend initiates a new
+ * connection. On the virtualized device removal frontend goes into the
+ * XenbusStateInitialising state.
+ *
+ * Note on XenbusStateReconfiguring state of the frontend: if backend has
+ * unrecoverable errors then frontend cannot send requests to the backend
+ * and thus cannot provide functionality of the virtualized device anymore.
+ * After backend is back to normal the virtualized device may still hold some
+ * state: configuration in use, allocated buffers, client application state etc.
+ * So, in most cases, this will require frontend to implement complex recovery
+ * reconnect logic. Instead, by going into XenbusStateReconfiguring state,
+ * frontend will make sure no new clients of the virtualized device are
+ * accepted, allow existing client(s) to exit gracefully by signaling error
+ * state etc.
+ * Once all the clients are gone frontend can reinitialize the virtualized
+ * device and get into XenbusStateInitialising state again signaling the
+ * backend that a new connection can be made.
+ *
+ * There are multiple conditions possible under which frontend will go from
+ * XenbusStateReconfiguring into XenbusStateInitialising, some of them are OS
+ * specific. For example:
+ * 1. The underlying OS framework may provide callbacks to signal that the last
+ * client of the virtualized device has gone and the device can be removed
+ * 2. Frontend can schedule a deferred work (timer/tasklet/workqueue)
+ * to periodically check if this is the right time to re-try removal of
+ * the virtualized device.
+ * 3. By any other means.
+ *
+ ******************************************************************************
+ * PCM FORMATS
+ ******************************************************************************
+ *
+ * XENSND_PCM_FORMAT_<format>[_<endian>]
+ *
+ * format: <S/U/F><bits> or <name>
+ * S - signed, U - unsigned, F - float
+ * bits - 8, 16, 24, 32
+ * name - MU_LAW, GSM, etc.
+ *
+ * endian: <LE/BE>, may be absent
+ * LE - Little endian, BE - Big endian
+ */
+#define XENSND_PCM_FORMAT_S8 0
+#define XENSND_PCM_FORMAT_U8 1
+#define XENSND_PCM_FORMAT_S16_LE 2
+#define XENSND_PCM_FORMAT_S16_BE 3
+#define XENSND_PCM_FORMAT_U16_LE 4
+#define XENSND_PCM_FORMAT_U16_BE 5
+#define XENSND_PCM_FORMAT_S24_LE 6
+#define XENSND_PCM_FORMAT_S24_BE 7
+#define XENSND_PCM_FORMAT_U24_LE 8
+#define XENSND_PCM_FORMAT_U24_BE 9
+#define XENSND_PCM_FORMAT_S32_LE 10
+#define XENSND_PCM_FORMAT_S32_BE 11
+#define XENSND_PCM_FORMAT_U32_LE 12
+#define XENSND_PCM_FORMAT_U32_BE 13
+#define XENSND_PCM_FORMAT_F32_LE 14 /* 4-byte float, IEEE-754 32-bit, */
+#define XENSND_PCM_FORMAT_F32_BE 15 /* range -1.0 to 1.0 */
+#define XENSND_PCM_FORMAT_F64_LE 16 /* 8-byte float, IEEE-754 64-bit, */
+#define XENSND_PCM_FORMAT_F64_BE 17 /* range -1.0 to 1.0 */
+#define XENSND_PCM_FORMAT_IEC958_SUBFRAME_LE 18
+#define XENSND_PCM_FORMAT_IEC958_SUBFRAME_BE 19
+#define XENSND_PCM_FORMAT_MU_LAW 20
+#define XENSND_PCM_FORMAT_A_LAW 21
+#define XENSND_PCM_FORMAT_IMA_ADPCM 22
+#define XENSND_PCM_FORMAT_MPEG 23
+#define XENSND_PCM_FORMAT_GSM 24
+
+/*
+ ******************************************************************************
+ * REQUEST CODES
+ ******************************************************************************
+ */
+#define XENSND_OP_OPEN 0
+#define XENSND_OP_CLOSE 1
+#define XENSND_OP_READ 2
+#define XENSND_OP_WRITE 3
+#define XENSND_OP_SET_VOLUME 4
+#define XENSND_OP_GET_VOLUME 5
+#define XENSND_OP_MUTE 6
+#define XENSND_OP_UNMUTE 7
+
+/*
+ ******************************************************************************
+ * XENSTORE FIELD AND PATH NAME STRINGS, HELPERS
+ ******************************************************************************
+ */
+#define XENSND_DRIVER_NAME "vsnd"
+
+#define XENSND_LIST_SEPARATOR ","
+/* Field names */
+#define XENSND_FIELD_BE_VERSIONS "versions"
+#define XENSND_FIELD_FE_VERSION "version"
+#define XENSND_FIELD_VCARD_SHORT_NAME "short-name"
+#define XENSND_FIELD_VCARD_LONG_NAME "long-name"
+#define XENSND_FIELD_RING_REF "ring-ref"
+#define XENSND_FIELD_EVT_CHNL "event-channel"
+#define XENSND_FIELD_DEVICE_NAME "name"
+#define XENSND_FIELD_TYPE "type"
+#define XENSND_FIELD_STREAM_UNIQUE_ID "unique-id"
+#define XENSND_FIELD_CHANNELS_MIN "channels-min"
+#define XENSND_FIELD_CHANNELS_MAX "channels-max"
+#define XENSND_FIELD_SAMPLE_RATES "sample-rates"
+#define XENSND_FIELD_SAMPLE_FORMATS "sample-formats"
+#define XENSND_FIELD_BUFFER_SIZE "buffer-size"
+
+/* Stream type field values. */
+#define XENSND_STREAM_TYPE_PLAYBACK "p"
+#define XENSND_STREAM_TYPE_CAPTURE "c"
+/* Sample rate max string length */
+#define XENSND_SAMPLE_RATE_MAX_LEN 11
+/* Sample format field values */
+#define XENSND_SAMPLE_FORMAT_MAX_LEN 24
+
+#define XENSND_PCM_FORMAT_S8_STR "s8"
+#define XENSND_PCM_FORMAT_U8_STR "u8"
+#define XENSND_PCM_FORMAT_S16_LE_STR "s16_le"
+#define XENSND_PCM_FORMAT_S16_BE_STR "s16_be"
+#define XENSND_PCM_FORMAT_U16_LE_STR "u16_le"
+#define XENSND_PCM_FORMAT_U16_BE_STR "u16_be"
+#define XENSND_PCM_FORMAT_S24_LE_STR "s24_le"
+#define XENSND_PCM_FORMAT_S24_BE_STR "s24_be"
+#define XENSND_PCM_FORMAT_U24_LE_STR "u24_le"
+#define XENSND_PCM_FORMAT_U24_BE_STR "u24_be"
+#define XENSND_PCM_FORMAT_S32_LE_STR "s32_le"
+#define XENSND_PCM_FORMAT_S32_BE_STR "s32_be"
+#define XENSND_PCM_FORMAT_U32_LE_STR "u32_le"
+#define XENSND_PCM_FORMAT_U32_BE_STR "u32_be"
+#define XENSND_PCM_FORMAT_F32_LE_STR "float_le"
+#define XENSND_PCM_FORMAT_F32_BE_STR "float_be"
+#define XENSND_PCM_FORMAT_F64_LE_STR "float64_le"
+#define XENSND_PCM_FORMAT_F64_BE_STR "float64_be"
+#define XENSND_PCM_FORMAT_IEC958_SUBFRAME_LE_STR "iec958_subframe_le"
+#define XENSND_PCM_FORMAT_IEC958_SUBFRAME_BE_STR "iec958_subframe_be"
+#define XENSND_PCM_FORMAT_MU_LAW_STR "mu_law"
+#define XENSND_PCM_FORMAT_A_LAW_STR "a_law"
+#define XENSND_PCM_FORMAT_IMA_ADPCM_STR "ima_adpcm"
+#define XENSND_PCM_FORMAT_MPEG_STR "mpeg"
+#define XENSND_PCM_FORMAT_GSM_STR "gsm"
+
+
+/*
+ ******************************************************************************
+ * STATUS RETURN CODES
+ ******************************************************************************
+ *
+ * Status return code is zero on success and -XEN_EXX on failure.
+ *
+ ******************************************************************************
+ * Assumptions
+ ******************************************************************************
+ * o usage of grant reference 0 as invalid grant reference:
+ * grant reference 0 is valid, but never exposed to a PV driver,
+ * because of the fact it is already in use/reserved by the PV console.
+ * o all references in this document to page sizes must be treated
+ * as pages of size XEN_PAGE_SIZE unless otherwise noted.
+ *
+ ******************************************************************************
+ * Description of the protocol between frontend and backend driver
+ ******************************************************************************
+ *
+ * The two halves of a Para-virtual sound driver communicate with
+ * each other using shared pages and event channels.
+ * Shared page contains a ring with request/response packets.
+ *
+ * Packets, used for input/output operations, e.g. read/write, set/get volume,
+ * etc., provide offset/length fields in order to allow asynchronous protocol
+ * operation with buffer space sharing: part of the buffer allocated at
+ * XENSND_OP_OPEN can be used for audio samples and part, for example,
+ * for volume control.
+ *
+ * All reserved fields in the structures below must be 0.
+ *
+ *---------------------------------- Requests ---------------------------------
+ *
+ * All request packets have the same length (32 octets)
+ * All request packets have common header:
+ * 0 1 2 3 octet
+ * +----------------+----------------+----------------+----------------+
+ * | id | operation | reserved | 4
+ * +----------------+----------------+----------------+----------------+
+ * | reserved | 8
+ * +----------------+----------------+----------------+----------------+
+ * id - uint16_t, private guest value, echoed in response
+ * operation - uint8_t, operation code, XENSND_OP_???
+ *
+ * For all packets which use offset and length:
+ * offset - uint32_t, read or write data offset within the shared buffer,
+ * passed with XENSND_OP_OPEN request, octets,
+ * [0; XENSND_OP_OPEN.buffer_sz - 1].
+ * length - uint32_t, read or write data length, octets
+ *
+ * Request open - open a PCM stream for playback or capture:
+ *
+ * 0 1 2 3 octet
+ * +----------------+----------------+----------------+----------------+
+ * | id | XENSND_OP_OPEN | reserved | 4
+ * +----------------+----------------+----------------+----------------+
+ * | reserved | 8
+ * +----------------+----------------+----------------+----------------+
+ * | pcm_rate | 12
+ * +----------------+----------------+----------------+----------------+
+ * | pcm_format | pcm_channels | reserved | 16
+ * +----------------+----------------+----------------+----------------+
+ * | buffer_sz | 20
+ * +----------------+----------------+----------------+----------------+
+ * | gref_directory | 24
+ * +----------------+----------------+----------------+----------------+
+ * | reserved | 28
+ * +----------------+----------------+----------------+----------------+
+ * |/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/|
+ * +----------------+----------------+----------------+----------------+
+ * | reserved | 32
+ * +----------------+----------------+----------------+----------------+
+ *
+ * pcm_rate - uint32_t, stream data rate, Hz
+ * pcm_format - uint8_t, XENSND_PCM_FORMAT_XXX value
+ * pcm_channels - uint8_t, number of channels of this stream,
+ * [channels-min; channels-max]
+ * buffer_sz - uint32_t, buffer size to be allocated, octets
+ * gref_directory - grant_ref_t, a reference to the first shared page
+ * describing shared buffer references. At least one page exists. If shared
+ * buffer size (buffer_sz) exceeds what can be addressed by this single page,
+ * then reference to the next page must be supplied (see gref_dir_next_page
+ * below)
+ */
+
+struct xensnd_open_req {
+ uint32_t pcm_rate;
+ uint8_t pcm_format;
+ uint8_t pcm_channels;
+ uint16_t reserved;
+ uint32_t buffer_sz;
+ grant_ref_t gref_directory;
+};
+
+/*
+ * Shared page for XENSND_OP_OPEN buffer descriptor (gref_directory in the
+ * request) employs a list of pages, describing all pages of the shared data
+ * buffer:
+ * 0 1 2 3 octet
+ * +----------------+----------------+----------------+----------------+
+ * | gref_dir_next_page | 4
+ * +----------------+----------------+----------------+----------------+
+ * | gref[0] | 8
+ * +----------------+----------------+----------------+----------------+
+ * |/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/|
+ * +----------------+----------------+----------------+----------------+
+ * | gref[i] | i*4+8
+ * +----------------+----------------+----------------+----------------+
+ * |/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/|
+ * +----------------+----------------+----------------+----------------+
+ * | gref[N - 1] | N*4+8
+ * +----------------+----------------+----------------+----------------+
+ *
+ * gref_dir_next_page - grant_ref_t, reference to the next page describing
+ * page directory. Must be 0 if there are no more pages in the list.
+ * gref[i] - grant_ref_t, reference to a shared page of the buffer
+ * allocated at XENSND_OP_OPEN
+ *
+ * Number of grant_ref_t entries in the whole page directory is not
+ * passed, but instead can be calculated as:
+ * num_grefs_total = (XENSND_OP_OPEN.buffer_sz + XEN_PAGE_SIZE - 1) /
+ * XEN_PAGE_SIZE
+ */
+
+struct xensnd_page_directory {
+ grant_ref_t gref_dir_next_page;
+ grant_ref_t gref[1]; /* Variable length */
+};
+
+/*
+ * Request close - close an opened pcm stream:
+ * 0 1 2 3 octet
+ * +----------------+----------------+----------------+----------------+
+ * | id | XENSND_OP_CLOSE| reserved | 4
+ * +----------------+----------------+----------------+----------------+
+ * | reserved | 8
+ * +----------------+----------------+----------------+----------------+
+ * |/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/|
+ * +----------------+----------------+----------------+----------------+
+ * | reserved | 32
+ * +----------------+----------------+----------------+----------------+
+ *
+ * Request read/write - used for read (for capture) or write (for playback):
+ * 0 1 2 3 octet
+ * +----------------+----------------+----------------+----------------+
+ * | id | operation | reserved | 4
+ * +----------------+----------------+----------------+----------------+
+ * | reserved | 8
+ * +----------------+----------------+----------------+----------------+
+ * | offset | 12
+ * +----------------+----------------+----------------+----------------+
+ * | length | 16
+ * +----------------+----------------+----------------+----------------+
+ * | reserved | 20
+ * +----------------+----------------+----------------+----------------+
+ * |/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/|
+ * +----------------+----------------+----------------+----------------+
+ * | reserved | 32
+ * +----------------+----------------+----------------+----------------+
+ *
+ * operation - XENSND_OP_READ for read or XENSND_OP_WRITE for write
+ */
+
+struct xensnd_rw_req {
+ uint32_t offset;
+ uint32_t length;
+};
+
+/*
+ * Request set/get volume - set/get channels' volume of the stream given:
+ * 0 1 2 3 octet
+ * +----------------+----------------+----------------+----------------+
+ * | id | operation | reserved | 4
+ * +----------------+----------------+----------------+----------------+
+ * | reserved | 8
+ * +----------------+----------------+----------------+----------------+
+ * | offset | 12
+ * +----------------+----------------+----------------+----------------+
+ * | length | 16
+ * +----------------+----------------+----------------+----------------+
+ * |/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/|
+ * +----------------+----------------+----------------+----------------+
+ * | reserved | 32
+ * +----------------+----------------+----------------+----------------+
+ *
+ * operation - XENSND_OP_SET_VOLUME for volume set
+ * or XENSND_OP_GET_VOLUME for volume get
+ * Buffer passed with XENSND_OP_OPEN is used to exchange volume
+ * values:
+ *
+ * 0 1 2 3 octet
+ * +----------------+----------------+----------------+----------------+
+ * | channel[0] | 4
+ * +----------------+----------------+----------------+----------------+
+ * |/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/|
+ * +----------------+----------------+----------------+----------------+
+ * | channel[i] | i*4
+ * +----------------+----------------+----------------+----------------+
+ * |/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/|
+ * +----------------+----------------+----------------+----------------+
+ * | channel[N - 1] | (N-1)*4
+ * +----------------+----------------+----------------+----------------+
+ *
+ * N = XENSND_OP_OPEN.pcm_channels
+ * i - uint8_t, index of a channel
+ * channel[i] - sint32_t, volume of i-th channel
+ * Volume is expressed as a signed value in steps of 0.001 dB,
+ * while 0 being 0 dB.
+ *
+ * Request mute/unmute - mute/unmute stream:
+ * 0 1 2 3 octet
+ * +----------------+----------------+----------------+----------------+
+ * | id | operation | reserved | 4
+ * +----------------+----------------+----------------+----------------+
+ * | reserved | 8
+ * +----------------+----------------+----------------+----------------+
+ * | offset | 12
+ * +----------------+----------------+----------------+----------------+
+ * | length | 16
+ * +----------------+----------------+----------------+----------------+
+ * |/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/|
+ * +----------------+----------------+----------------+----------------+
+ * | reserved | 32
+ * +----------------+----------------+----------------+----------------+
+ *
+ * operation - XENSND_OP_MUTE for mute or XENSND_OP_UNMUTE for unmute
+ * Buffer passed with XENSND_OP_OPEN is used to exchange mute/unmute
+ * values:
+ *
+ * 0 octet
+ * +----------------+----------------+----------------+----------------+
+ * | channel[0] | 4
+ * +----------------+----------------+----------------+----------------+
+ * +/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/|
+ * +----------------+----------------+----------------+----------------+
+ * | channel[i] | i*4
+ * +----------------+----------------+----------------+----------------+
+ * +/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/|
+ * +----------------+----------------+----------------+----------------+
+ * | channel[N - 1] | (N-1)*4
+ * +----------------+----------------+----------------+----------------+
+ *
+ * N = XENSND_OP_OPEN.pcm_channels
+ * i - uint8_t, index of a channel
+ * channel[i] - uint8_t, non-zero if i-th channel needs to be muted/unmuted
+ *
+ *------------------------------------ N.B. -----------------------------------
+ *
+ * The 'struct xensnd_rw_req' is also used for XENSND_OP_SET_VOLUME,
+ * XENSND_OP_GET_VOLUME, XENSND_OP_MUTE, XENSND_OP_UNMUTE.
+ */
+
+/*
+ *---------------------------------- Responses --------------------------------
+ *
+ * All response packets have the same length (32 octets)
+ *
+ * Response for all requests:
+ * 0 1 2 3 octet
+ * +----------------+----------------+----------------+----------------+
+ * | id | operation | reserved | 4
+ * +----------------+----------------+----------------+----------------+
+ * | status | 8
+ * +----------------+----------------+----------------+----------------+
+ * | reserved | 12
+ * +----------------+----------------+----------------+----------------+
+ * |/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/|
+ * +----------------+----------------+----------------+----------------+
+ * | reserved | 32
+ * +----------------+----------------+----------------+----------------+
+ *
+ * id - uint16_t, copied from the request
+ * operation - uint8_t, XENSND_OP_* - copied from request
+ * status - int32_t, response status, zero on success and -XEN_EXX on failure
+ */
+
+struct xensnd_req {
+ uint16_t id;
+ uint8_t operation;
+ uint8_t reserved[5];
+ union {
+ struct xensnd_open_req open;
+ struct xensnd_rw_req rw;
+ uint8_t reserved[24];
+ } op;
+};
+
+struct xensnd_resp {
+ uint16_t id;
+ uint8_t operation;
+ uint8_t reserved;
+ int32_t status;
+ uint8_t reserved1[24];
+};
+
+DEFINE_RING_TYPES(xen_sndif, struct xensnd_req, struct xensnd_resp);
+
+#endif /* __XEN_PUBLIC_IO_SNDIF_H__ */
+
+/*
+ * Local variables:
+ * mode: C
+ * c-file-style: "BSD"
+ * c-basic-offset: 4
+ * tab-width: 4
+ * indent-tabs-mode: nil
+ * End:
+ */
--
2.7.4
^ permalink raw reply related [flat|nested] 41+ messages in thread
* [For Linux 3/4] xen/sndif: add sound-device ABI
2017-04-07 8:30 [For Linux 0/4] PV protocol headers for Linux Kernel Oleksandr Andrushchenko
` (3 preceding siblings ...)
2017-04-07 8:30 ` Oleksandr Andrushchenko
@ 2017-04-07 8:30 ` Oleksandr Andrushchenko
2017-04-07 8:30 ` Oleksandr Andrushchenko
` (4 subsequent siblings)
9 siblings, 0 replies; 41+ messages in thread
From: Oleksandr Andrushchenko @ 2017-04-07 8:30 UTC (permalink / raw)
To: xen-devel
Cc: lars.kurth, Iurii Konovalenko, sstabellini, vlad.babchuk,
linux-kernel, Oleksandr Dmytryshyn, Oleksandr Andrushchenko,
julien.grall, andrii.anisov, olekstysh, andr2000, al1img,
Oleksandr Grytsov, joculator
From: Oleksandr Andrushchenko <oleksandr_andrushchenko@epam.com>
Add ABI for the two halves of a para-virtualized
sound driver to communicate with each other.
The ABI allows implementing audio playback and capture as
well as volume control and possibility to mute/unmute
audio sources.
Note: depending on the use-case backend can expose more sound
cards and PCM devices/streams than the underlying HW physically
has by employing SW mixers, configuring virtual sound streams,
channels etc. Thus, allowing fine tunned configurations per
frontend.
Reviewed-by: Konrad Rzeszutek Wilk <konrad.wilk@oracle.com>
Signed-off-by: Oleksandr Andrushchenko <oleksandr_andrushchenko@epam.com>
Signed-off-by: Oleksandr Grytsov <oleksandr_grytsov@epam.com>
Signed-off-by: Oleksandr Dmytryshyn <oleksandr.dmytryshyn@globallogic.com>
Signed-off-by: Iurii Konovalenko <iurii.konovalenko@globallogic.com>
---
include/xen/interface/io/sndif.h | 803 +++++++++++++++++++++++++++++++++++++++
1 file changed, 803 insertions(+)
create mode 100644 include/xen/interface/io/sndif.h
diff --git a/include/xen/interface/io/sndif.h b/include/xen/interface/io/sndif.h
new file mode 100644
index 000000000000..ea9b81024cd9
--- /dev/null
+++ b/include/xen/interface/io/sndif.h
@@ -0,0 +1,803 @@
+/******************************************************************************
+ * sndif.h
+ *
+ * Unified sound-device I/O interface for Xen guest OSes.
+ *
+ * Permission is hereby granted, free of charge, to any person obtaining a copy
+ * of this software and associated documentation files (the "Software"), to
+ * deal in the Software without restriction, including without limitation the
+ * rights to use, copy, modify, merge, publish, distribute, sublicense, and/or
+ * sell copies of the Software, and to permit persons to whom the Software is
+ * furnished to do so, subject to the following conditions:
+ *
+ * The above copyright notice and this permission notice shall be included in
+ * all copies or substantial portions of the Software.
+ *
+ * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
+ * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
+ * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE
+ * AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER
+ * LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING
+ * FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER
+ * DEALINGS IN THE SOFTWARE.
+ *
+ * Copyright (C) 2013-2015 GlobalLogic Inc.
+ * Copyright (C) 2016-2017 EPAM Systems Inc.
+ *
+ * Authors: Oleksandr Andrushchenko <oleksandr_andrushchenko@epam.com>
+ * Oleksandr Grytsov <oleksandr_grytsov@epam.com>
+ * Oleksandr Dmytryshyn <oleksandr.dmytryshyn@globallogic.com>
+ * Iurii Konovalenko <iurii.konovalenko@globallogic.com>
+ */
+
+#ifndef __XEN_PUBLIC_IO_SNDIF_H__
+#define __XEN_PUBLIC_IO_SNDIF_H__
+
+#include "ring.h"
+#include "../grant_table.h"
+
+/*
+ ******************************************************************************
+ * Feature and Parameter Negotiation
+ ******************************************************************************
+ *
+ * Front->back notifications: when enqueuing a new request, sending a
+ * notification can be made conditional on xensnd_req (i.e., the generic
+ * hold-off mechanism provided by the ring macros). Backends must set
+ * xensnd_req appropriately (e.g., using RING_FINAL_CHECK_FOR_REQUESTS()).
+ *
+ * Back->front notifications: when enqueuing a new response, sending a
+ * notification can be made conditional on xensnd_resp (i.e., the generic
+ * hold-off mechanism provided by the ring macros). Frontends must set
+ * xensnd_resp appropriately (e.g., using RING_FINAL_CHECK_FOR_RESPONSES()).
+ *
+ * The two halves of a para-virtual sound card driver utilize nodes within
+ * XenStore to communicate capabilities and to negotiate operating parameters.
+ * This section enumerates these nodes which reside in the respective front and
+ * backend portions of XenStore, following the XenBus convention.
+ *
+ * All data in XenStore is stored as strings. Nodes specifying numeric
+ * values are encoded in decimal. Integer value ranges listed below are
+ * expressed as fixed sized integer types capable of storing the conversion
+ * of a properly formated node string, without loss of information.
+ *
+ ******************************************************************************
+ * Example configuration
+ ******************************************************************************
+ *
+ * Note: depending on the use-case backend can expose more sound cards and
+ * PCM devices/streams than the underlying HW physically has by employing
+ * SW mixers, configuring virtual sound streams, channels etc.
+ *
+ * This is an example of backend and frontend configuration:
+ *
+ *--------------------------------- Backend -----------------------------------
+ *
+ * /local/domain/0/backend/vsnd/1/0/frontend-id = "1"
+ * /local/domain/0/backend/vsnd/1/0/frontend = "/local/domain/1/device/vsnd/0"
+ * /local/domain/0/backend/vsnd/1/0/state = "4"
+ * /local/domain/0/backend/vsnd/1/0/versions = "1,2"
+ *
+ *--------------------------------- Frontend ----------------------------------
+ *
+ * /local/domain/1/device/vsnd/0/backend-id = "0"
+ * /local/domain/1/device/vsnd/0/backend = "/local/domain/0/backend/vsnd/1/0"
+ * /local/domain/1/device/vsnd/0/state = "4"
+ * /local/domain/1/device/vsnd/0/version = "1"
+ *
+ *----------------------------- Card configuration ----------------------------
+ *
+ * /local/domain/1/device/vsnd/0/short-name = "Card short name"
+ * /local/domain/1/device/vsnd/0/long-name = "Card long name"
+ * /local/domain/1/device/vsnd/0/sample-rates = "8000,32000,44100,48000,96000"
+ * /local/domain/1/device/vsnd/0/sample-formats = "s8,u8,s16_le,s16_be"
+ * /local/domain/1/device/vsnd/0/buffer-size = "262144"
+ *
+ *------------------------------- PCM device 0 --------------------------------
+ *
+ * /local/domain/1/device/vsnd/0/0/name = "General analog"
+ * /local/domain/1/device/vsnd/0/0/channels-max = "5"
+ *
+ *----------------------------- Stream 0, playback ----------------------------
+ *
+ * /local/domain/1/device/vsnd/0/0/0/type = "p"
+ * /local/domain/1/device/vsnd/0/0/0/sample-formats = "s8,u8"
+ * /local/domain/1/device/vsnd/0/0/0/unique-id = "0"
+ *
+ * /local/domain/1/device/vsnd/0/0/0/ring-ref = "386"
+ * /local/domain/1/device/vsnd/0/0/0/event-channel = "15"
+ *
+ *------------------------------ Stream 1, capture ----------------------------
+ *
+ * /local/domain/1/device/vsnd/0/0/1/type = "c"
+ * /local/domain/1/device/vsnd/0/0/1/channels-max = "2"
+ * /local/domain/1/device/vsnd/0/0/1/unique-id = "1"
+ *
+ * /local/domain/1/device/vsnd/0/0/1/ring-ref = "384"
+ * /local/domain/1/device/vsnd/0/0/1/event-channel = "13"
+ *
+ *------------------------------- PCM device 1 --------------------------------
+ *
+ * /local/domain/1/device/vsnd/0/1/name = "HDMI-0"
+ * /local/domain/1/device/vsnd/0/1/sample-rates = "8000,32000,44100"
+ *
+ *------------------------------ Stream 0, capture ----------------------------
+ *
+ * /local/domain/1/device/vsnd/0/1/0/type = "c"
+ * /local/domain/1/device/vsnd/0/1/0/unique-id = "2"
+ *
+ * /local/domain/1/device/vsnd/0/1/0/ring-ref = "387"
+ * /local/domain/1/device/vsnd/0/1/0/event-channel = "151"
+ *
+ *------------------------------- PCM device 2 --------------------------------
+ *
+ * /local/domain/1/device/vsnd/0/2/name = "SPDIF"
+ *
+ *----------------------------- Stream 0, playback ----------------------------
+ *
+ * /local/domain/1/device/vsnd/0/2/0/type = "p"
+ * /local/domain/1/device/vsnd/0/2/0/unique-id = "3"
+ *
+ * /local/domain/1/device/vsnd/0/2/0/ring-ref = "389"
+ * /local/domain/1/device/vsnd/0/2/0/event-channel = "152"
+ *
+ ******************************************************************************
+ * Backend XenBus Nodes
+ ******************************************************************************
+ *
+ *----------------------------- Protocol version ------------------------------
+ *
+ * versions
+ * Values: <string>
+ *
+ * List of XENSND_LIST_SEPARATOR separated protocol versions supported
+ * by the backend. For example "1,2,3".
+ *
+ ******************************************************************************
+ * Frontend XenBus Nodes
+ ******************************************************************************
+ *
+ *-------------------------------- Addressing ---------------------------------
+ *
+ * dom-id
+ * Values: <uint16_t>
+ *
+ * Domain identifier.
+ *
+ * dev-id
+ * Values: <uint16_t>
+ *
+ * Device identifier.
+ *
+ * pcm-dev-idx
+ * Values: <uint8_t>
+ *
+ * Zero based contigous index of the PCM device.
+ *
+ * stream-idx
+ * Values: <uint8_t>
+ *
+ * Zero based contigous index of the stream of the PCM device.
+ *
+ * The following pattern is used for addressing:
+ * /local/domain/<dom-id>/device/vsnd/<dev-id>/<pcm-dev-idx>/<stream-idx>/...
+ *
+ *----------------------------- Protocol version ------------------------------
+ *
+ * version
+ * Values: <string>
+ *
+ * Protocol version, chosen among the ones supported by the backend.
+ *
+ *------------------------------- PCM settings --------------------------------
+ *
+ * Every virtualized sound frontend has a set of PCM devices and streams, each
+ * could be individually configured. Part of the PCM configuration can be
+ * defined at higher level of the hierarchy and be fully or partially re-used
+ * by the underlying layers. These configuration values are:
+ * o number of channels (min/max)
+ * o supported sample rates
+ * o supported sample formats.
+ * E.g. one can define these values for the whole card, device or stream.
+ * Every underlying layer in turn can re-define some or all of them to better
+ * fit its needs. For example, card may define number of channels to be
+ * in [1; 8] range, and some particular stream may be limited to [1; 2] only.
+ * The rule is that the underlying layer must be a subset of the upper layer
+ * range.
+ *
+ * channels-min
+ * Values: <uint8_t>
+ *
+ * The minimum amount of channels that is supported, [1; channels-max].
+ * Optional, if not set or omitted a value of 1 is used.
+ *
+ * channels-max
+ * Values: <uint8_t>
+ *
+ * The maximum amount of channels that is supported.
+ * Must be at least <channels-min>.
+ *
+ * sample-rates
+ * Values: <list of uint32_t>
+ *
+ * List of supported sample rates separated by XENSND_LIST_SEPARATOR.
+ * Sample rates are expressed as a list of decimal values w/o any
+ * ordering requirement.
+ *
+ * sample-formats
+ * Values: <list of XENSND_PCM_FORMAT_XXX_STR>
+ *
+ * List of supported sample formats separated by XENSND_LIST_SEPARATOR.
+ * Items must not exceed XENSND_SAMPLE_FORMAT_MAX_LEN length.
+ *
+ * buffer-size
+ * Values: <uint32_t>
+ *
+ * The maximum size in octets of the buffer to allocate per stream.
+ *
+ *----------------------- Virtual sound card settings -------------------------
+ * short-name
+ * Values: <char[32]>
+ *
+ * Short name of the virtual sound card. Optional.
+ *
+ * long-name
+ * Values: <char[80]>
+ *
+ * Long name of the virtual sound card. Optional.
+ *
+ *----------------------------- Device settings -------------------------------
+ * name
+ * Values: <char[80]>
+ *
+ * Name of the sound device within the virtual sound card. Optional.
+ *
+ *----------------------------- Stream settings -------------------------------
+ *
+ * type
+ * Values: "p", "c"
+ *
+ * Stream type: "p" - playback stream, "c" - capture stream
+ *
+ * If both capture and playback are needed then two streams need to be
+ * defined under the same device.
+ *
+ * unique-id
+ * Values: <uint32_t>
+ *
+ * After stream initialization it is assigned a unique ID (within the front
+ * driver), so every stream of the frontend can be identified by the
+ * backend by this ID. This is not equal to stream-idx as the later is
+ * zero based within the device, but this index is contigous within the
+ * driver.
+ *
+ *-------------------- Stream Request Transport Parameters --------------------
+ *
+ * event-channel
+ * Values: <uint32_t>
+ *
+ * The identifier of the Xen event channel used to signal activity
+ * in the ring buffer.
+ *
+ * ring-ref
+ * Values: <uint32_t>
+ *
+ * The Xen grant reference granting permission for the backend to map
+ * a sole page in a single page sized ring buffer.
+ *
+ ******************************************************************************
+ * STATE DIAGRAMS
+ ******************************************************************************
+ *
+ * Tool stack creates front and back state nodes with initial state
+ * XenbusStateInitialising.
+ * Tool stack creates and sets up frontend sound configuration nodes per domain.
+ *
+ * Front Back
+ * ================================= =====================================
+ * XenbusStateInitialising XenbusStateInitialising
+ * o Query backend device identification
+ * data.
+ * o Open and validate backend device.
+ * |
+ * |
+ * V
+ * XenbusStateInitWait
+ *
+ * o Query frontend configuration
+ * o Allocate and initialize
+ * event channels per configured
+ * playback/capture stream.
+ * o Publish transport parameters
+ * that will be in effect during
+ * this connection.
+ * |
+ * |
+ * V
+ * XenbusStateInitialised
+ *
+ * o Query frontend transport parameters.
+ * o Connect to the event channels.
+ * |
+ * |
+ * V
+ * XenbusStateConnected
+ *
+ * o Create and initialize OS
+ * virtual sound device instances
+ * as per configuration.
+ * |
+ * |
+ * V
+ * XenbusStateConnected
+ *
+ * XenbusStateUnknown
+ * XenbusStateClosed
+ * XenbusStateClosing
+ * o Remove virtual sound device
+ * o Remove event channels
+ * |
+ * |
+ * V
+ * XenbusStateClosed
+ *
+ *------------------------------- Recovery flow -------------------------------
+ *
+ * In case of frontend unrecoverable errors backend handles that as
+ * if frontend goes into the XenbusStateClosed state.
+ *
+ * In case of backend unrecoverable errors frontend tries removing
+ * the virtualized device. If this is possible at the moment of error,
+ * then frontend goes into the XenbusStateInitialising state and is ready for
+ * new connection with backend. If the virtualized device is still in use and
+ * cannot be removed, then frontend goes into the XenbusStateReconfiguring state
+ * until either the virtualized device removed or backend initiates a new
+ * connection. On the virtualized device removal frontend goes into the
+ * XenbusStateInitialising state.
+ *
+ * Note on XenbusStateReconfiguring state of the frontend: if backend has
+ * unrecoverable errors then frontend cannot send requests to the backend
+ * and thus cannot provide functionality of the virtualized device anymore.
+ * After backend is back to normal the virtualized device may still hold some
+ * state: configuration in use, allocated buffers, client application state etc.
+ * So, in most cases, this will require frontend to implement complex recovery
+ * reconnect logic. Instead, by going into XenbusStateReconfiguring state,
+ * frontend will make sure no new clients of the virtualized device are
+ * accepted, allow existing client(s) to exit gracefully by signaling error
+ * state etc.
+ * Once all the clients are gone frontend can reinitialize the virtualized
+ * device and get into XenbusStateInitialising state again signaling the
+ * backend that a new connection can be made.
+ *
+ * There are multiple conditions possible under which frontend will go from
+ * XenbusStateReconfiguring into XenbusStateInitialising, some of them are OS
+ * specific. For example:
+ * 1. The underlying OS framework may provide callbacks to signal that the last
+ * client of the virtualized device has gone and the device can be removed
+ * 2. Frontend can schedule a deferred work (timer/tasklet/workqueue)
+ * to periodically check if this is the right time to re-try removal of
+ * the virtualized device.
+ * 3. By any other means.
+ *
+ ******************************************************************************
+ * PCM FORMATS
+ ******************************************************************************
+ *
+ * XENSND_PCM_FORMAT_<format>[_<endian>]
+ *
+ * format: <S/U/F><bits> or <name>
+ * S - signed, U - unsigned, F - float
+ * bits - 8, 16, 24, 32
+ * name - MU_LAW, GSM, etc.
+ *
+ * endian: <LE/BE>, may be absent
+ * LE - Little endian, BE - Big endian
+ */
+#define XENSND_PCM_FORMAT_S8 0
+#define XENSND_PCM_FORMAT_U8 1
+#define XENSND_PCM_FORMAT_S16_LE 2
+#define XENSND_PCM_FORMAT_S16_BE 3
+#define XENSND_PCM_FORMAT_U16_LE 4
+#define XENSND_PCM_FORMAT_U16_BE 5
+#define XENSND_PCM_FORMAT_S24_LE 6
+#define XENSND_PCM_FORMAT_S24_BE 7
+#define XENSND_PCM_FORMAT_U24_LE 8
+#define XENSND_PCM_FORMAT_U24_BE 9
+#define XENSND_PCM_FORMAT_S32_LE 10
+#define XENSND_PCM_FORMAT_S32_BE 11
+#define XENSND_PCM_FORMAT_U32_LE 12
+#define XENSND_PCM_FORMAT_U32_BE 13
+#define XENSND_PCM_FORMAT_F32_LE 14 /* 4-byte float, IEEE-754 32-bit, */
+#define XENSND_PCM_FORMAT_F32_BE 15 /* range -1.0 to 1.0 */
+#define XENSND_PCM_FORMAT_F64_LE 16 /* 8-byte float, IEEE-754 64-bit, */
+#define XENSND_PCM_FORMAT_F64_BE 17 /* range -1.0 to 1.0 */
+#define XENSND_PCM_FORMAT_IEC958_SUBFRAME_LE 18
+#define XENSND_PCM_FORMAT_IEC958_SUBFRAME_BE 19
+#define XENSND_PCM_FORMAT_MU_LAW 20
+#define XENSND_PCM_FORMAT_A_LAW 21
+#define XENSND_PCM_FORMAT_IMA_ADPCM 22
+#define XENSND_PCM_FORMAT_MPEG 23
+#define XENSND_PCM_FORMAT_GSM 24
+
+/*
+ ******************************************************************************
+ * REQUEST CODES
+ ******************************************************************************
+ */
+#define XENSND_OP_OPEN 0
+#define XENSND_OP_CLOSE 1
+#define XENSND_OP_READ 2
+#define XENSND_OP_WRITE 3
+#define XENSND_OP_SET_VOLUME 4
+#define XENSND_OP_GET_VOLUME 5
+#define XENSND_OP_MUTE 6
+#define XENSND_OP_UNMUTE 7
+
+/*
+ ******************************************************************************
+ * XENSTORE FIELD AND PATH NAME STRINGS, HELPERS
+ ******************************************************************************
+ */
+#define XENSND_DRIVER_NAME "vsnd"
+
+#define XENSND_LIST_SEPARATOR ","
+/* Field names */
+#define XENSND_FIELD_BE_VERSIONS "versions"
+#define XENSND_FIELD_FE_VERSION "version"
+#define XENSND_FIELD_VCARD_SHORT_NAME "short-name"
+#define XENSND_FIELD_VCARD_LONG_NAME "long-name"
+#define XENSND_FIELD_RING_REF "ring-ref"
+#define XENSND_FIELD_EVT_CHNL "event-channel"
+#define XENSND_FIELD_DEVICE_NAME "name"
+#define XENSND_FIELD_TYPE "type"
+#define XENSND_FIELD_STREAM_UNIQUE_ID "unique-id"
+#define XENSND_FIELD_CHANNELS_MIN "channels-min"
+#define XENSND_FIELD_CHANNELS_MAX "channels-max"
+#define XENSND_FIELD_SAMPLE_RATES "sample-rates"
+#define XENSND_FIELD_SAMPLE_FORMATS "sample-formats"
+#define XENSND_FIELD_BUFFER_SIZE "buffer-size"
+
+/* Stream type field values. */
+#define XENSND_STREAM_TYPE_PLAYBACK "p"
+#define XENSND_STREAM_TYPE_CAPTURE "c"
+/* Sample rate max string length */
+#define XENSND_SAMPLE_RATE_MAX_LEN 11
+/* Sample format field values */
+#define XENSND_SAMPLE_FORMAT_MAX_LEN 24
+
+#define XENSND_PCM_FORMAT_S8_STR "s8"
+#define XENSND_PCM_FORMAT_U8_STR "u8"
+#define XENSND_PCM_FORMAT_S16_LE_STR "s16_le"
+#define XENSND_PCM_FORMAT_S16_BE_STR "s16_be"
+#define XENSND_PCM_FORMAT_U16_LE_STR "u16_le"
+#define XENSND_PCM_FORMAT_U16_BE_STR "u16_be"
+#define XENSND_PCM_FORMAT_S24_LE_STR "s24_le"
+#define XENSND_PCM_FORMAT_S24_BE_STR "s24_be"
+#define XENSND_PCM_FORMAT_U24_LE_STR "u24_le"
+#define XENSND_PCM_FORMAT_U24_BE_STR "u24_be"
+#define XENSND_PCM_FORMAT_S32_LE_STR "s32_le"
+#define XENSND_PCM_FORMAT_S32_BE_STR "s32_be"
+#define XENSND_PCM_FORMAT_U32_LE_STR "u32_le"
+#define XENSND_PCM_FORMAT_U32_BE_STR "u32_be"
+#define XENSND_PCM_FORMAT_F32_LE_STR "float_le"
+#define XENSND_PCM_FORMAT_F32_BE_STR "float_be"
+#define XENSND_PCM_FORMAT_F64_LE_STR "float64_le"
+#define XENSND_PCM_FORMAT_F64_BE_STR "float64_be"
+#define XENSND_PCM_FORMAT_IEC958_SUBFRAME_LE_STR "iec958_subframe_le"
+#define XENSND_PCM_FORMAT_IEC958_SUBFRAME_BE_STR "iec958_subframe_be"
+#define XENSND_PCM_FORMAT_MU_LAW_STR "mu_law"
+#define XENSND_PCM_FORMAT_A_LAW_STR "a_law"
+#define XENSND_PCM_FORMAT_IMA_ADPCM_STR "ima_adpcm"
+#define XENSND_PCM_FORMAT_MPEG_STR "mpeg"
+#define XENSND_PCM_FORMAT_GSM_STR "gsm"
+
+
+/*
+ ******************************************************************************
+ * STATUS RETURN CODES
+ ******************************************************************************
+ *
+ * Status return code is zero on success and -XEN_EXX on failure.
+ *
+ ******************************************************************************
+ * Assumptions
+ ******************************************************************************
+ * o usage of grant reference 0 as invalid grant reference:
+ * grant reference 0 is valid, but never exposed to a PV driver,
+ * because of the fact it is already in use/reserved by the PV console.
+ * o all references in this document to page sizes must be treated
+ * as pages of size XEN_PAGE_SIZE unless otherwise noted.
+ *
+ ******************************************************************************
+ * Description of the protocol between frontend and backend driver
+ ******************************************************************************
+ *
+ * The two halves of a Para-virtual sound driver communicate with
+ * each other using shared pages and event channels.
+ * Shared page contains a ring with request/response packets.
+ *
+ * Packets, used for input/output operations, e.g. read/write, set/get volume,
+ * etc., provide offset/length fields in order to allow asynchronous protocol
+ * operation with buffer space sharing: part of the buffer allocated at
+ * XENSND_OP_OPEN can be used for audio samples and part, for example,
+ * for volume control.
+ *
+ * All reserved fields in the structures below must be 0.
+ *
+ *---------------------------------- Requests ---------------------------------
+ *
+ * All request packets have the same length (32 octets)
+ * All request packets have common header:
+ * 0 1 2 3 octet
+ * +----------------+----------------+----------------+----------------+
+ * | id | operation | reserved | 4
+ * +----------------+----------------+----------------+----------------+
+ * | reserved | 8
+ * +----------------+----------------+----------------+----------------+
+ * id - uint16_t, private guest value, echoed in response
+ * operation - uint8_t, operation code, XENSND_OP_???
+ *
+ * For all packets which use offset and length:
+ * offset - uint32_t, read or write data offset within the shared buffer,
+ * passed with XENSND_OP_OPEN request, octets,
+ * [0; XENSND_OP_OPEN.buffer_sz - 1].
+ * length - uint32_t, read or write data length, octets
+ *
+ * Request open - open a PCM stream for playback or capture:
+ *
+ * 0 1 2 3 octet
+ * +----------------+----------------+----------------+----------------+
+ * | id | XENSND_OP_OPEN | reserved | 4
+ * +----------------+----------------+----------------+----------------+
+ * | reserved | 8
+ * +----------------+----------------+----------------+----------------+
+ * | pcm_rate | 12
+ * +----------------+----------------+----------------+----------------+
+ * | pcm_format | pcm_channels | reserved | 16
+ * +----------------+----------------+----------------+----------------+
+ * | buffer_sz | 20
+ * +----------------+----------------+----------------+----------------+
+ * | gref_directory | 24
+ * +----------------+----------------+----------------+----------------+
+ * | reserved | 28
+ * +----------------+----------------+----------------+----------------+
+ * |/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/|
+ * +----------------+----------------+----------------+----------------+
+ * | reserved | 32
+ * +----------------+----------------+----------------+----------------+
+ *
+ * pcm_rate - uint32_t, stream data rate, Hz
+ * pcm_format - uint8_t, XENSND_PCM_FORMAT_XXX value
+ * pcm_channels - uint8_t, number of channels of this stream,
+ * [channels-min; channels-max]
+ * buffer_sz - uint32_t, buffer size to be allocated, octets
+ * gref_directory - grant_ref_t, a reference to the first shared page
+ * describing shared buffer references. At least one page exists. If shared
+ * buffer size (buffer_sz) exceeds what can be addressed by this single page,
+ * then reference to the next page must be supplied (see gref_dir_next_page
+ * below)
+ */
+
+struct xensnd_open_req {
+ uint32_t pcm_rate;
+ uint8_t pcm_format;
+ uint8_t pcm_channels;
+ uint16_t reserved;
+ uint32_t buffer_sz;
+ grant_ref_t gref_directory;
+};
+
+/*
+ * Shared page for XENSND_OP_OPEN buffer descriptor (gref_directory in the
+ * request) employs a list of pages, describing all pages of the shared data
+ * buffer:
+ * 0 1 2 3 octet
+ * +----------------+----------------+----------------+----------------+
+ * | gref_dir_next_page | 4
+ * +----------------+----------------+----------------+----------------+
+ * | gref[0] | 8
+ * +----------------+----------------+----------------+----------------+
+ * |/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/|
+ * +----------------+----------------+----------------+----------------+
+ * | gref[i] | i*4+8
+ * +----------------+----------------+----------------+----------------+
+ * |/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/|
+ * +----------------+----------------+----------------+----------------+
+ * | gref[N - 1] | N*4+8
+ * +----------------+----------------+----------------+----------------+
+ *
+ * gref_dir_next_page - grant_ref_t, reference to the next page describing
+ * page directory. Must be 0 if there are no more pages in the list.
+ * gref[i] - grant_ref_t, reference to a shared page of the buffer
+ * allocated at XENSND_OP_OPEN
+ *
+ * Number of grant_ref_t entries in the whole page directory is not
+ * passed, but instead can be calculated as:
+ * num_grefs_total = (XENSND_OP_OPEN.buffer_sz + XEN_PAGE_SIZE - 1) /
+ * XEN_PAGE_SIZE
+ */
+
+struct xensnd_page_directory {
+ grant_ref_t gref_dir_next_page;
+ grant_ref_t gref[1]; /* Variable length */
+};
+
+/*
+ * Request close - close an opened pcm stream:
+ * 0 1 2 3 octet
+ * +----------------+----------------+----------------+----------------+
+ * | id | XENSND_OP_CLOSE| reserved | 4
+ * +----------------+----------------+----------------+----------------+
+ * | reserved | 8
+ * +----------------+----------------+----------------+----------------+
+ * |/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/|
+ * +----------------+----------------+----------------+----------------+
+ * | reserved | 32
+ * +----------------+----------------+----------------+----------------+
+ *
+ * Request read/write - used for read (for capture) or write (for playback):
+ * 0 1 2 3 octet
+ * +----------------+----------------+----------------+----------------+
+ * | id | operation | reserved | 4
+ * +----------------+----------------+----------------+----------------+
+ * | reserved | 8
+ * +----------------+----------------+----------------+----------------+
+ * | offset | 12
+ * +----------------+----------------+----------------+----------------+
+ * | length | 16
+ * +----------------+----------------+----------------+----------------+
+ * | reserved | 20
+ * +----------------+----------------+----------------+----------------+
+ * |/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/|
+ * +----------------+----------------+----------------+----------------+
+ * | reserved | 32
+ * +----------------+----------------+----------------+----------------+
+ *
+ * operation - XENSND_OP_READ for read or XENSND_OP_WRITE for write
+ */
+
+struct xensnd_rw_req {
+ uint32_t offset;
+ uint32_t length;
+};
+
+/*
+ * Request set/get volume - set/get channels' volume of the stream given:
+ * 0 1 2 3 octet
+ * +----------------+----------------+----------------+----------------+
+ * | id | operation | reserved | 4
+ * +----------------+----------------+----------------+----------------+
+ * | reserved | 8
+ * +----------------+----------------+----------------+----------------+
+ * | offset | 12
+ * +----------------+----------------+----------------+----------------+
+ * | length | 16
+ * +----------------+----------------+----------------+----------------+
+ * |/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/|
+ * +----------------+----------------+----------------+----------------+
+ * | reserved | 32
+ * +----------------+----------------+----------------+----------------+
+ *
+ * operation - XENSND_OP_SET_VOLUME for volume set
+ * or XENSND_OP_GET_VOLUME for volume get
+ * Buffer passed with XENSND_OP_OPEN is used to exchange volume
+ * values:
+ *
+ * 0 1 2 3 octet
+ * +----------------+----------------+----------------+----------------+
+ * | channel[0] | 4
+ * +----------------+----------------+----------------+----------------+
+ * |/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/|
+ * +----------------+----------------+----------------+----------------+
+ * | channel[i] | i*4
+ * +----------------+----------------+----------------+----------------+
+ * |/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/|
+ * +----------------+----------------+----------------+----------------+
+ * | channel[N - 1] | (N-1)*4
+ * +----------------+----------------+----------------+----------------+
+ *
+ * N = XENSND_OP_OPEN.pcm_channels
+ * i - uint8_t, index of a channel
+ * channel[i] - sint32_t, volume of i-th channel
+ * Volume is expressed as a signed value in steps of 0.001 dB,
+ * while 0 being 0 dB.
+ *
+ * Request mute/unmute - mute/unmute stream:
+ * 0 1 2 3 octet
+ * +----------------+----------------+----------------+----------------+
+ * | id | operation | reserved | 4
+ * +----------------+----------------+----------------+----------------+
+ * | reserved | 8
+ * +----------------+----------------+----------------+----------------+
+ * | offset | 12
+ * +----------------+----------------+----------------+----------------+
+ * | length | 16
+ * +----------------+----------------+----------------+----------------+
+ * |/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/|
+ * +----------------+----------------+----------------+----------------+
+ * | reserved | 32
+ * +----------------+----------------+----------------+----------------+
+ *
+ * operation - XENSND_OP_MUTE for mute or XENSND_OP_UNMUTE for unmute
+ * Buffer passed with XENSND_OP_OPEN is used to exchange mute/unmute
+ * values:
+ *
+ * 0 octet
+ * +----------------+----------------+----------------+----------------+
+ * | channel[0] | 4
+ * +----------------+----------------+----------------+----------------+
+ * +/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/|
+ * +----------------+----------------+----------------+----------------+
+ * | channel[i] | i*4
+ * +----------------+----------------+----------------+----------------+
+ * +/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/|
+ * +----------------+----------------+----------------+----------------+
+ * | channel[N - 1] | (N-1)*4
+ * +----------------+----------------+----------------+----------------+
+ *
+ * N = XENSND_OP_OPEN.pcm_channels
+ * i - uint8_t, index of a channel
+ * channel[i] - uint8_t, non-zero if i-th channel needs to be muted/unmuted
+ *
+ *------------------------------------ N.B. -----------------------------------
+ *
+ * The 'struct xensnd_rw_req' is also used for XENSND_OP_SET_VOLUME,
+ * XENSND_OP_GET_VOLUME, XENSND_OP_MUTE, XENSND_OP_UNMUTE.
+ */
+
+/*
+ *---------------------------------- Responses --------------------------------
+ *
+ * All response packets have the same length (32 octets)
+ *
+ * Response for all requests:
+ * 0 1 2 3 octet
+ * +----------------+----------------+----------------+----------------+
+ * | id | operation | reserved | 4
+ * +----------------+----------------+----------------+----------------+
+ * | status | 8
+ * +----------------+----------------+----------------+----------------+
+ * | reserved | 12
+ * +----------------+----------------+----------------+----------------+
+ * |/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/|
+ * +----------------+----------------+----------------+----------------+
+ * | reserved | 32
+ * +----------------+----------------+----------------+----------------+
+ *
+ * id - uint16_t, copied from the request
+ * operation - uint8_t, XENSND_OP_* - copied from request
+ * status - int32_t, response status, zero on success and -XEN_EXX on failure
+ */
+
+struct xensnd_req {
+ uint16_t id;
+ uint8_t operation;
+ uint8_t reserved[5];
+ union {
+ struct xensnd_open_req open;
+ struct xensnd_rw_req rw;
+ uint8_t reserved[24];
+ } op;
+};
+
+struct xensnd_resp {
+ uint16_t id;
+ uint8_t operation;
+ uint8_t reserved;
+ int32_t status;
+ uint8_t reserved1[24];
+};
+
+DEFINE_RING_TYPES(xen_sndif, struct xensnd_req, struct xensnd_resp);
+
+#endif /* __XEN_PUBLIC_IO_SNDIF_H__ */
+
+/*
+ * Local variables:
+ * mode: C
+ * c-file-style: "BSD"
+ * c-basic-offset: 4
+ * tab-width: 4
+ * indent-tabs-mode: nil
+ * End:
+ */
--
2.7.4
_______________________________________________
Xen-devel mailing list
Xen-devel@lists.xen.org
https://lists.xen.org/xen-devel
^ permalink raw reply related [flat|nested] 41+ messages in thread
* [For Linux 4/4] xen/displif: add ABI for para-virtual display
2017-04-07 8:30 [For Linux 0/4] PV protocol headers for Linux Kernel Oleksandr Andrushchenko
` (5 preceding siblings ...)
2017-04-07 8:30 ` Oleksandr Andrushchenko
@ 2017-04-07 8:30 ` Oleksandr Andrushchenko
2017-04-07 11:31 ` [Xen-devel] " Juergen Gross
` (3 more replies)
2017-04-07 8:30 ` Oleksandr Andrushchenko
` (2 subsequent siblings)
9 siblings, 4 replies; 41+ messages in thread
From: Oleksandr Andrushchenko @ 2017-04-07 8:30 UTC (permalink / raw)
To: xen-devel
Cc: joculator, al1img, vlad.babchuk, andrii.anisov, olekstysh,
konrad.wilk, lars.kurth, julien.grall, andr2000, sstabellini,
linux-kernel, Oleksandr Andrushchenko, Oleksandr Grytsov
From: Oleksandr Andrushchenko <oleksandr_andrushchenko@epam.com>
This is the ABI for the two halves of a para-virtualized
display driver.
This protocol aims to provide a unified protocol which fits more
sophisticated use-cases than a framebuffer device can handle. At the
moment basic functionality is supported with the intention to extend:
o multiple dynamically allocated/destroyed framebuffers
o buffers of arbitrary sizes
o better configuration options including multiple display support
Note: existing fbif can be used together with displif running at the
same time, e.g. on Linux one provides framebuffer and another DRM/KMS
Future extensions to the existing protocol may include:
o allow display/connector cloning
o allow allocating objects other than display buffers
o add planes/overlays support
o support scaling
o support rotation
Note, that this protocol doesn't use ring macros for
bi-directional exchange (PV calls/9pfs) bacause:
o it statically defines the use of a single page
for the ring buffer
o it uses direct memory access to ring's contents
w/o memory copying
o re-uses the same idea that kbdif/fbif use
which for this use-case seems to be appropriate
==================================================
Rationale for introducing this protocol instead of
using the existing fbif:
==================================================
1. In/out event sizes
o fbif - 40 octets
o displif - 40 octets
This is only the initial version of the displif protocol
which means that there could be requests which will not fit
(WRT introducing some GPU related functionality
later on). In that case we cannot alter fbif sizes as we need to
be backward compatible an will be forced to handle those
apart of fbif.
2. Shared page
Displif doesn't use anything like struct xenfb_page, but
DEFINE_RING_TYPES(xen_displif, struct xendispl_req, struct
xendispl_resp) which is a better and more common way.
Output events use a shared page which only has in_cons and in_prod
and all the rest is used for incoming events. Here struct xenfb_page
could probably be used as is despite the fact that it only has a half
of a page for incoming events which is only 50 events. (consider
something like 60Hz display)
3. Amount of changes.
fbif only provides XENFB_TYPE_UPDATE and XENFB_TYPE_RESIZE
events, so it looks like it is easier to get fb support into displif
than vice versa. displif at the moment has 6 requests and 1 event,
multiple connector support, etc.
Reviewed-by: Konrad Rzeszutek Wilk <konrad.wilk@oracle.com>
Signed-off-by: Oleksandr Grytsov <oleksandr_grytsov@epam.com>
Signed-off-by: Oleksandr Andrushchenko <oleksandr_andrushchenko@epam.com>
---
include/xen/interface/io/displif.h | 864 +++++++++++++++++++++++++++++++++++++
1 file changed, 864 insertions(+)
create mode 100644 include/xen/interface/io/displif.h
diff --git a/include/xen/interface/io/displif.h b/include/xen/interface/io/displif.h
new file mode 100644
index 000000000000..ac2b489ecf27
--- /dev/null
+++ b/include/xen/interface/io/displif.h
@@ -0,0 +1,864 @@
+/******************************************************************************
+ * displif.h
+ *
+ * Unified display device I/O interface for Xen guest OSes.
+ *
+ * Permission is hereby granted, free of charge, to any person obtaining a copy
+ * of this software and associated documentation files (the "Software"), to
+ * deal in the Software without restriction, including without limitation the
+ * rights to use, copy, modify, merge, publish, distribute, sublicense, and/or
+ * sell copies of the Software, and to permit persons to whom the Software is
+ * furnished to do so, subject to the following conditions:
+ *
+ * The above copyright notice and this permission notice shall be included in
+ * all copies or substantial portions of the Software.
+ *
+ * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
+ * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
+ * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE
+ * AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER
+ * LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING
+ * FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER
+ * DEALINGS IN THE SOFTWARE.
+ *
+ * Copyright (C) 2016-2017 EPAM Systems Inc.
+ *
+ * Authors: Oleksandr Andrushchenko <oleksandr_andrushchenko@epam.com>
+ * Oleksandr Grytsov <oleksandr_grytsov@epam.com>
+ */
+
+#ifndef __XEN_PUBLIC_IO_DISPLIF_H__
+#define __XEN_PUBLIC_IO_DISPLIF_H__
+
+#include "ring.h"
+#include "../grant_table.h"
+
+/*
+ ******************************************************************************
+ * Protocol version
+ ******************************************************************************
+ */
+#define XENDISPL_PROTOCOL_VERSION "1"
+
+/*
+ ******************************************************************************
+ * Main features provided by the protocol
+ ******************************************************************************
+ * This protocol aims to provide a unified protocol which fits more
+ * sophisticated use-cases than a framebuffer device can handle. At the
+ * moment basic functionality is supported with the intention to be extended:
+ * o multiple dynamically allocated/destroyed framebuffers
+ * o buffers of arbitrary sizes
+ * o buffer allocation at either back or front end
+ * o better configuration options including multiple display support
+ *
+ * Note: existing fbif can be used together with displif running at the
+ * same time, e.g. on Linux one provides framebuffer and another DRM/KMS
+ *
+ * Note: display resolution (XenStore's "resolution" property) defines
+ * visible area of the virtual display. At the same time resolution of
+ * the display and frame buffers may differ: buffers can be smaller, equal
+ * or bigger than the visible area. This is to enable use-cases, where backend
+ * may do some post-processing of the display and frame buffers supplied,
+ * e.g. those buffers can be just a part of the final composition.
+ *
+ ******************************************************************************
+ * Direction of improvements
+ ******************************************************************************
+ * Future extensions to the existing protocol may include:
+ * o display/connector cloning
+ * o allocation of objects other than display buffers
+ * o plane/overlay support
+ * o scaling support
+ * o rotation support
+ *
+ ******************************************************************************
+ * Feature and Parameter Negotiation
+ ******************************************************************************
+ *
+ * Front->back notifications: when enqueuing a new request, sending a
+ * notification can be made conditional on xendispl_req (i.e., the generic
+ * hold-off mechanism provided by the ring macros). Backends must set
+ * xendispl_req appropriately (e.g., using RING_FINAL_CHECK_FOR_REQUESTS()).
+ *
+ * Back->front notifications: when enqueuing a new response, sending a
+ * notification can be made conditional on xendispl_resp (i.e., the generic
+ * hold-off mechanism provided by the ring macros). Frontends must set
+ * xendispl_resp appropriately (e.g., using RING_FINAL_CHECK_FOR_RESPONSES()).
+ *
+ * The two halves of a para-virtual display driver utilize nodes within
+ * XenStore to communicate capabilities and to negotiate operating parameters.
+ * This section enumerates these nodes which reside in the respective front and
+ * backend portions of XenStore, following the XenBus convention.
+ *
+ * All data in XenStore is stored as strings. Nodes specifying numeric
+ * values are encoded in decimal. Integer value ranges listed below are
+ * expressed as fixed sized integer types capable of storing the conversion
+ * of a properly formated node string, without loss of information.
+ *
+ ******************************************************************************
+ * Example configuration
+ ******************************************************************************
+ *
+ * Note: depending on the use-case backend can expose more display connectors
+ * than the underlying HW physically has by employing SW graphics compositors
+ *
+ * This is an example of backend and frontend configuration:
+ *
+ *--------------------------------- Backend -----------------------------------
+ *
+ * /local/domain/0/backend/vdispl/1/0/frontend-id = "1"
+ * /local/domain/0/backend/vdispl/1/0/frontend = "/local/domain/1/device/vdispl/0"
+ * /local/domain/0/backend/vdispl/1/0/state = "4"
+ * /local/domain/0/backend/vdispl/1/0/versions = "1,2"
+ *
+ *--------------------------------- Frontend ----------------------------------
+ *
+ * /local/domain/1/device/vdispl/0/backend-id = "0"
+ * /local/domain/1/device/vdispl/0/backend = "/local/domain/0/backend/vdispl/1/0"
+ * /local/domain/1/device/vdispl/0/state = "4"
+ * /local/domain/1/device/vdispl/0/version = "1"
+ * /local/domain/1/device/vdispl/0/be-alloc = "1"
+ *
+ *-------------------------- Connector 0 configuration ------------------------
+ *
+ * /local/domain/1/device/vdispl/0/0/resolution = "1920x1080"
+ * /local/domain/1/device/vdispl/0/0/req-ring-ref = "2832"
+ * /local/domain/1/device/vdispl/0/0/req-event-channel = "15"
+ * /local/domain/1/device/vdispl/0/0/evt-ring-ref = "387"
+ * /local/domain/1/device/vdispl/0/0/evt-event-channel = "16"
+ *
+ *-------------------------- Connector 1 configuration ------------------------
+ *
+ * /local/domain/1/device/vdispl/0/1/resolution = "800x600"
+ * /local/domain/1/device/vdispl/0/1/req-ring-ref = "2833"
+ * /local/domain/1/device/vdispl/0/1/req-event-channel = "17"
+ * /local/domain/1/device/vdispl/0/1/evt-ring-ref = "388"
+ * /local/domain/1/device/vdispl/0/1/evt-event-channel = "18"
+ *
+ ******************************************************************************
+ * Backend XenBus Nodes
+ ******************************************************************************
+ *
+ *----------------------------- Protocol version ------------------------------
+ *
+ * versions
+ * Values: <string>
+ *
+ * List of XENDISPL_LIST_SEPARATOR separated protocol versions supported
+ * by the backend. For example "1,2,3".
+ *
+ ******************************************************************************
+ * Frontend XenBus Nodes
+ ******************************************************************************
+ *
+ *-------------------------------- Addressing ---------------------------------
+ *
+ * dom-id
+ * Values: <uint16_t>
+ *
+ * Domain identifier.
+ *
+ * dev-id
+ * Values: <uint16_t>
+ *
+ * Device identifier.
+ *
+ * conn-idx
+ * Values: <uint8_t>
+ *
+ * Zero based contigous index of the connector.
+ * /local/domain/<dom-id>/device/vdispl/<dev-id>/<conn-idx>/...
+ *
+ *----------------------------- Protocol version ------------------------------
+ *
+ * version
+ * Values: <string>
+ *
+ * Protocol version, chosen among the ones supported by the backend.
+ *
+ *------------------------- Backend buffer allocation -------------------------
+ *
+ * be-alloc
+ * Values: "0", "1"
+ *
+ * If value is set to "1", then backend can be a buffer provider/allocator
+ * for this domain during XENDISPL_OP_DBUF_CREATE operation (see below
+ * for negotiation).
+ * If value is not "1" or omitted frontend must allocate buffers itself.
+ *
+ *----------------------------- Connector settings ----------------------------
+ *
+ * resolution
+ * Values: <width, uint32_t>x<height, uint32_t>
+ *
+ * Width and height of the connector in pixels separated by
+ * XENDISPL_RESOLUTION_SEPARATOR. This defines visible area of the
+ * display.
+ *
+ *------------------ Connector Request Transport Parameters -------------------
+ *
+ * This communication path is used to deliver requests from frontend to backend
+ * and get the corresponding responses from backend to frontend,
+ * set up per connector.
+ *
+ * req-event-channel
+ * Values: <uint32_t>
+ *
+ * The identifier of the Xen connector's control event channel
+ * used to signal activity in the ring buffer.
+ *
+ * req-ring-ref
+ * Values: <uint32_t>
+ *
+ * The Xen grant reference granting permission for the backend to map
+ * a sole page of connector's control ring buffer.
+ *
+ *------------------- Connector Event Transport Parameters --------------------
+ *
+ * This communication path is used to deliver asynchronous events from backend
+ * to frontend, set up per connector.
+ *
+ * evt-event-channel
+ * Values: <uint32_t>
+ *
+ * The identifier of the Xen connector's event channel
+ * used to signal activity in the ring buffer.
+ *
+ * evt-ring-ref
+ * Values: <uint32_t>
+ *
+ * The Xen grant reference granting permission for the backend to map
+ * a sole page of connector's event ring buffer.
+ */
+
+/*
+ ******************************************************************************
+ * STATE DIAGRAMS
+ ******************************************************************************
+ *
+ * Tool stack creates front and back state nodes with initial state
+ * XenbusStateInitialising.
+ * Tool stack creates and sets up frontend display configuration
+ * nodes per domain.
+ *
+ *-------------------------------- Normal flow --------------------------------
+ *
+ * Front Back
+ * ================================= =====================================
+ * XenbusStateInitialising XenbusStateInitialising
+ * o Query backend device identification
+ * data.
+ * o Open and validate backend device.
+ * |
+ * |
+ * V
+ * XenbusStateInitWait
+ *
+ * o Query frontend configuration
+ * o Allocate and initialize
+ * event channels per configured
+ * connector.
+ * o Publish transport parameters
+ * that will be in effect during
+ * this connection.
+ * |
+ * |
+ * V
+ * XenbusStateInitialised
+ *
+ * o Query frontend transport parameters.
+ * o Connect to the event channels.
+ * |
+ * |
+ * V
+ * XenbusStateConnected
+ *
+ * o Create and initialize OS
+ * virtual display connectors
+ * as per configuration.
+ * |
+ * |
+ * V
+ * XenbusStateConnected
+ *
+ * XenbusStateUnknown
+ * XenbusStateClosed
+ * XenbusStateClosing
+ * o Remove virtual display device
+ * o Remove event channels
+ * |
+ * |
+ * V
+ * XenbusStateClosed
+ *
+ *------------------------------- Recovery flow -------------------------------
+ *
+ * In case of frontend unrecoverable errors backend handles that as
+ * if frontend goes into the XenbusStateClosed state.
+ *
+ * In case of backend unrecoverable errors frontend tries removing
+ * the virtualized device. If this is possible at the moment of error,
+ * then frontend goes into the XenbusStateInitialising state and is ready for
+ * new connection with backend. If the virtualized device is still in use and
+ * cannot be removed, then frontend goes into the XenbusStateReconfiguring state
+ * until either the virtualized device is removed or backend initiates a new
+ * connection. On the virtualized device removal frontend goes into the
+ * XenbusStateInitialising state.
+ *
+ * Note on XenbusStateReconfiguring state of the frontend: if backend has
+ * unrecoverable errors then frontend cannot send requests to the backend
+ * and thus cannot provide functionality of the virtualized device anymore.
+ * After backend is back to normal the virtualized device may still hold some
+ * state: configuration in use, allocated buffers, client application state etc.
+ * In most cases, this will require frontend to implement complex recovery
+ * reconnect logic. Instead, by going into XenbusStateReconfiguring state,
+ * frontend will make sure no new clients of the virtualized device are
+ * accepted, allow existing client(s) to exit gracefully by signaling error
+ * state etc.
+ * Once all the clients are gone frontend can reinitialize the virtualized
+ * device and get into XenbusStateInitialising state again signaling the
+ * backend that a new connection can be made.
+ *
+ * There are multiple conditions possible under which frontend will go from
+ * XenbusStateReconfiguring into XenbusStateInitialising, some of them are OS
+ * specific. For example:
+ * 1. The underlying OS framework may provide callbacks to signal that the last
+ * client of the virtualized device has gone and the device can be removed
+ * 2. Frontend can schedule a deferred work (timer/tasklet/workqueue)
+ * to periodically check if this is the right time to re-try removal of
+ * the virtualized device.
+ * 3. By any other means.
+ *
+ ******************************************************************************
+ * REQUEST CODES
+ ******************************************************************************
+ * Request codes [0; 15] are reserved and must not be used
+ */
+
+#define XENDISPL_OP_DBUF_CREATE 0x10
+#define XENDISPL_OP_DBUF_DESTROY 0x11
+#define XENDISPL_OP_FB_ATTACH 0x12
+#define XENDISPL_OP_FB_DETACH 0x13
+#define XENDISPL_OP_SET_CONFIG 0x14
+#define XENDISPL_OP_PG_FLIP 0x15
+
+/*
+ ******************************************************************************
+ * EVENT CODES
+ ******************************************************************************
+ */
+#define XENDISPL_EVT_PG_FLIP 0x00
+
+/*
+ ******************************************************************************
+ * XENSTORE FIELD AND PATH NAME STRINGS, HELPERS
+ ******************************************************************************
+ */
+#define XENDISPL_DRIVER_NAME "vdispl"
+
+#define XENDISPL_LIST_SEPARATOR ","
+#define XENDISPL_RESOLUTION_SEPARATOR "x"
+
+#define XENDISPL_FIELD_BE_VERSIONS "versions"
+#define XENDISPL_FIELD_FE_VERSION "version"
+#define XENDISPL_FIELD_REQ_RING_REF "req-ring-ref"
+#define XENDISPL_FIELD_REQ_CHANNEL "req-event-channel"
+#define XENDISPL_FIELD_EVT_RING_REF "evt-ring-ref"
+#define XENDISPL_FIELD_EVT_CHANNEL "evt-event-channel"
+#define XENDISPL_FIELD_RESOLUTION "resolution"
+#define XENDISPL_FIELD_BE_ALLOC "be-alloc"
+
+/*
+ ******************************************************************************
+ * STATUS RETURN CODES
+ ******************************************************************************
+ *
+ * Status return code is zero on success and -XEN_EXX on failure.
+ *
+ ******************************************************************************
+ * Assumptions
+ ******************************************************************************
+ * o usage of grant reference 0 as invalid grant reference:
+ * grant reference 0 is valid, but never exposed to a PV driver,
+ * because of the fact it is already in use/reserved by the PV console.
+ * o all references in this document to page sizes must be treated
+ * as pages of size XEN_PAGE_SIZE unless otherwise noted.
+ *
+ ******************************************************************************
+ * Description of the protocol between frontend and backend driver
+ ******************************************************************************
+ *
+ * The two halves of a Para-virtual display driver communicate with
+ * each other using shared pages and event channels.
+ * Shared page contains a ring with request/response packets.
+ *
+ * All reserved fields in the structures below must be 0.
+ * Display buffers's cookie of value 0 is treated as invalid.
+ * Framebuffer's cookie of value 0 is treated as invalid.
+ *
+ * For all request/response/event packets that use cookies:
+ * dbuf_cookie - uint64_t, unique to guest domain value used by the backend
+ * to map remote display buffer to its local one
+ * fb_cookie - uint64_t, unique to guest domain value used by the backend
+ * to map remote framebuffer to its local one
+ *
+ *---------------------------------- Requests ---------------------------------
+ *
+ * All requests/responses, which are not connector specific, must be sent over
+ * control ring of the connector which has the index value of 0:
+ * /local/domain/<dom-id>/device/vdispl/<dev-id>/0/req-ring-ref
+ *
+ * All request packets have the same length (64 octets)
+ * All request packets have common header:
+ * 0 1 2 3 octet
+ * +----------------+----------------+----------------+----------------+
+ * | id | operation | reserved | 4
+ * +----------------+----------------+----------------+----------------+
+ * | reserved | 8
+ * +----------------+----------------+----------------+----------------+
+ * id - uint16_t, private guest value, echoed in response
+ * operation - uint8_t, operation code, XENDISPL_OP_???
+ *
+ * Request dbuf creation - request creation of a display buffer.
+ * 0 1 2 3 octet
+ * +----------------+----------------+----------------+----------------+
+ * | id |_OP_DBUF_CREATE | reserved | 4
+ * +----------------+----------------+----------------+----------------+
+ * | reserved | 8
+ * +----------------+----------------+----------------+----------------+
+ * | dbuf_cookie low 32-bit | 12
+ * +----------------+----------------+----------------+----------------+
+ * | dbuf_cookie high 32-bit | 16
+ * +----------------+----------------+----------------+----------------+
+ * | width | 20
+ * +----------------+----------------+----------------+----------------+
+ * | height | 24
+ * +----------------+----------------+----------------+----------------+
+ * | bpp | 28
+ * +----------------+----------------+----------------+----------------+
+ * | buffer_sz | 32
+ * +----------------+----------------+----------------+----------------+
+ * | flags | 36
+ * +----------------+----------------+----------------+----------------+
+ * | gref_directory | 40
+ * +----------------+----------------+----------------+----------------+
+ * | reserved | 44
+ * +----------------+----------------+----------------+----------------+
+ * |/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/|
+ * +----------------+----------------+----------------+----------------+
+ * | reserved | 64
+ * +----------------+----------------+----------------+----------------+
+ *
+ * Must be sent over control ring of the connector which has the index
+ * value of 0:
+ * /local/domain/<dom-id>/device/vdispl/<dev-id>/0/req-ring-ref
+ * All unused bits in flags field must be set to 0.
+ *
+ * An attempt to create multiple display buffers with the same dbuf_cookie is
+ * an error. dbuf_cookie can be re-used after destroying the corresponding
+ * display buffer.
+ *
+ * Width and height of the display buffers can be smaller, equal or bigger
+ * than the connector's resolution. Depth/pixel format of the individual
+ * buffers can differ as well.
+ *
+ * width - uint32_t, width in pixels
+ * height - uint32_t, height in pixels
+ * bpp - uint32_t, bits per pixel
+ * buffer_sz - uint32_t, buffer size to be allocated, octets
+ * flags - uint32_t, flags of the operation
+ * o XENDISPL_DBUF_FLG_REQ_ALLOC - if set, then backend is requested
+ * to allocate the buffer with the parameters provided in this request.
+ * Page directory is handled as follows:
+ * Frontend on request:
+ * o allocates pages for the directory (gref_directory,
+ * gref_dir_next_page(s)
+ * o grants permissions for the pages of the directory to the backend
+ * o sets gref_dir_next_page fields
+ * Backend on response:
+ * o grants permissions for the pages of the buffer allocated to
+ * the frontend
+ * o fills in page directory with grant references
+ * (gref[] in struct xendispl_page_directory)
+ * gref_directory - grant_ref_t, a reference to the first shared page
+ * describing shared buffer references. At least one page exists. If shared
+ * buffer size (buffer_sz) exceeds what can be addressed by this single page,
+ * then reference to the next page must be supplied (see gref_dir_next_page
+ * below)
+ */
+
+#define XENDISPL_DBUF_FLG_REQ_ALLOC (1 << 0)
+
+struct xendispl_dbuf_create_req {
+ uint64_t dbuf_cookie;
+ uint32_t width;
+ uint32_t height;
+ uint32_t bpp;
+ uint32_t buffer_sz;
+ uint32_t flags;
+ grant_ref_t gref_directory;
+};
+
+/*
+ * Shared page for XENDISPL_OP_DBUF_CREATE buffer descriptor (gref_directory in
+ * the request) employs a list of pages, describing all pages of the shared
+ * data buffer:
+ * 0 1 2 3 octet
+ * +----------------+----------------+----------------+----------------+
+ * | gref_dir_next_page | 4
+ * +----------------+----------------+----------------+----------------+
+ * | gref[0] | 8
+ * +----------------+----------------+----------------+----------------+
+ * |/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/|
+ * +----------------+----------------+----------------+----------------+
+ * | gref[i] | i*4+8
+ * +----------------+----------------+----------------+----------------+
+ * |/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/|
+ * +----------------+----------------+----------------+----------------+
+ * | gref[N - 1] | N*4+8
+ * +----------------+----------------+----------------+----------------+
+ *
+ * gref_dir_next_page - grant_ref_t, reference to the next page describing
+ * page directory. Must be 0 if there are no more pages in the list.
+ * gref[i] - grant_ref_t, reference to a shared page of the buffer
+ * allocated at XENDISPL_OP_DBUF_CREATE
+ *
+ * Number of grant_ref_t entries in the whole page directory is not
+ * passed, but instead can be calculated as:
+ * num_grefs_total = (XENDISPL_OP_DBUF_CREATE.buffer_sz + XEN_PAGE_SIZE - 1) /
+ * XEN_PAGE_SIZE
+ */
+
+struct xendispl_page_directory {
+ grant_ref_t gref_dir_next_page;
+ grant_ref_t gref[1]; /* Variable length */
+};
+
+/*
+ * Request dbuf destruction - destroy a previously allocated display buffer:
+ * 0 1 2 3 octet
+ * +----------------+----------------+----------------+----------------+
+ * | id |_OP_DBUF_DESTROY| reserved | 4
+ * +----------------+----------------+----------------+----------------+
+ * | reserved | 8
+ * +----------------+----------------+----------------+----------------+
+ * | dbuf_cookie low 32-bit | 12
+ * +----------------+----------------+----------------+----------------+
+ * | dbuf_cookie high 32-bit | 16
+ * +----------------+----------------+----------------+----------------+
+ * | reserved | 20
+ * +----------------+----------------+----------------+----------------+
+ * |/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/|
+ * +----------------+----------------+----------------+----------------+
+ * | reserved | 64
+ * +----------------+----------------+----------------+----------------+
+ *
+ * Must be sent over control ring of the connector which has the index
+ * value of 0:
+ * /local/domain/<dom-id>/device/vdispl/<dev-id>/0/req-ring-ref
+ */
+
+struct xendispl_dbuf_destroy_req {
+ uint64_t dbuf_cookie;
+};
+
+/*
+ * Request framebuffer attachment - request attachment of a framebuffer to
+ * previously created display buffer.
+ * 0 1 2 3 octet
+ * +----------------+----------------+----------------+----------------+
+ * | id | _OP_FB_ATTACH | reserved | 4
+ * +----------------+----------------+----------------+----------------+
+ * | reserved | 8
+ * +----------------+----------------+----------------+----------------+
+ * | dbuf_cookie low 32-bit | 12
+ * +----------------+----------------+----------------+----------------+
+ * | dbuf_cookie high 32-bit | 16
+ * +----------------+----------------+----------------+----------------+
+ * | fb_cookie low 32-bit | 20
+ * +----------------+----------------+----------------+----------------+
+ * | fb_cookie high 32-bit | 24
+ * +----------------+----------------+----------------+----------------+
+ * | width | 28
+ * +----------------+----------------+----------------+----------------+
+ * | height | 32
+ * +----------------+----------------+----------------+----------------+
+ * | pixel_format | 36
+ * +----------------+----------------+----------------+----------------+
+ * | reserved | 40
+ * +----------------+----------------+----------------+----------------+
+ * |/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/|
+ * +----------------+----------------+----------------+----------------+
+ * | reserved | 64
+ * +----------------+----------------+----------------+----------------+
+ *
+ * Must be sent over control ring of the connector which has the index
+ * value of 0:
+ * /local/domain/<dom-id>/device/vdispl/<dev-id>/0/req-ring-ref
+ * Width and height can be smaller, equal or bigger than the connector's
+ * resolution.
+ *
+ * An attempt to create multiple frame buffers with the same fb_cookie is
+ * an error. fb_cookie can be re-used after destroying the corresponding
+ * frame buffer.
+ *
+ * width - uint32_t, width in pixels
+ * height - uint32_t, height in pixels
+ * pixel_format - uint32_t, pixel format of the framebuffer, FOURCC code
+ */
+
+struct xendispl_fb_attach_req {
+ uint64_t dbuf_cookie;
+ uint64_t fb_cookie;
+ uint32_t width;
+ uint32_t height;
+ uint32_t pixel_format;
+};
+
+/*
+ * Request framebuffer detach - detach a previously
+ * attached framebuffer from the display buffer in request:
+ * 0 1 2 3 octet
+ * +----------------+----------------+----------------+----------------+
+ * | id | _OP_FB_DETACH | reserved | 4
+ * +----------------+----------------+----------------+----------------+
+ * | reserved | 8
+ * +----------------+----------------+----------------+----------------+
+ * | fb_cookie low 32-bit | 12
+ * +----------------+----------------+----------------+----------------+
+ * | fb_cookie high 32-bit | 16
+ * +----------------+----------------+----------------+----------------+
+ * | reserved | 20
+ * +----------------+----------------+----------------+----------------+
+ * |/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/|
+ * +----------------+----------------+----------------+----------------+
+ * | reserved | 64
+ * +----------------+----------------+----------------+----------------+
+ *
+ * Must be sent over control ring of the connector which has the index
+ * value of 0:
+ * /local/domain/<dom-id>/device/vdispl/<dev-id>/0/req-ring-ref
+ */
+
+struct xendispl_fb_detach_req {
+ uint64_t fb_cookie;
+};
+
+/*
+ * Request configuration set/reset - request to set or reset
+ * the configuration/mode of the display:
+ * 0 1 2 3 octet
+ * +----------------+----------------+----------------+----------------+
+ * | id | _OP_SET_CONFIG | reserved | 4
+ * +----------------+----------------+----------------+----------------+
+ * | reserved | 8
+ * +----------------+----------------+----------------+----------------+
+ * | fb_cookie low 32-bit | 12
+ * +----------------+----------------+----------------+----------------+
+ * | fb_cookie high 32-bit | 16
+ * +----------------+----------------+----------------+----------------+
+ * | x | 20
+ * +----------------+----------------+----------------+----------------+
+ * | y | 24
+ * +----------------+----------------+----------------+----------------+
+ * | width | 28
+ * +----------------+----------------+----------------+----------------+
+ * | height | 32
+ * +----------------+----------------+----------------+----------------+
+ * | bpp | 40
+ * +----------------+----------------+----------------+----------------+
+ * | reserved | 44
+ * +----------------+----------------+----------------+----------------+
+ * |/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/|
+ * +----------------+----------------+----------------+----------------+
+ * | reserved | 64
+ * +----------------+----------------+----------------+----------------+
+ *
+ * Pass all zeros to reset, otherwise command is treated as
+ * configuration set.
+ * Framebuffer's cookie defines which framebuffer/dbuf must be
+ * displayed while enabling display (applying configuration).
+ * x, y, width and height are bound by the connector's resolution and must not
+ * exceed it.
+ *
+ * x - uint32_t, starting position in pixels by X axis
+ * y - uint32_t, starting position in pixels by Y axis
+ * width - uint32_t, width in pixels
+ * height - uint32_t, height in pixels
+ * bpp - uint32_t, bits per pixel
+ */
+
+struct xendispl_set_config_req {
+ uint64_t fb_cookie;
+ uint32_t x;
+ uint32_t y;
+ uint32_t width;
+ uint32_t height;
+ uint32_t bpp;
+};
+
+/*
+ * Request page flip - request to flip a page identified by the framebuffer
+ * cookie:
+ * 0 1 2 3 octet
+ * +----------------+----------------+----------------+----------------+
+ * | id | _OP_PG_FLIP | reserved | 4
+ * +----------------+----------------+----------------+----------------+
+ * | reserved | 8
+ * +----------------+----------------+----------------+----------------+
+ * | fb_cookie low 32-bit | 12
+ * +----------------+----------------+----------------+----------------+
+ * | fb_cookie high 32-bit | 16
+ * +----------------+----------------+----------------+----------------+
+ * | reserved | 20
+ * +----------------+----------------+----------------+----------------+
+ * |/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/|
+ * +----------------+----------------+----------------+----------------+
+ * | reserved | 64
+ * +----------------+----------------+----------------+----------------+
+ */
+
+struct xendispl_page_flip_req {
+ uint64_t fb_cookie;
+};
+
+/*
+ *---------------------------------- Responses --------------------------------
+ *
+ * All response packets have the same length (64 octets)
+ *
+ * All response packets have common header:
+ * 0 1 2 3 octet
+ * +----------------+----------------+----------------+----------------+
+ * | id | reserved | 4
+ * +----------------+----------------+----------------+----------------+
+ * | status | 8
+ * +----------------+----------------+----------------+----------------+
+ * | reserved | 12
+ * +----------------+----------------+----------------+----------------+
+ * |/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/|
+ * +----------------+----------------+----------------+----------------+
+ * | reserved | 64
+ * +----------------+----------------+----------------+----------------+
+ *
+ * id - uint16_t, private guest value, echoed from request
+ * status - int32_t, response status, zero on success and -XEN_EXX on failure
+ *
+ *----------------------------------- Events ----------------------------------
+ *
+ * Events are sent via a shared page allocated by the front and propagated by
+ * evt-event-channel/evt-ring-ref XenStore entries
+ * All event packets have the same length (64 octets)
+ * All event packets have common header:
+ * 0 1 2 3 octet
+ * +----------------+----------------+----------------+----------------+
+ * | id | type | reserved | 4
+ * +----------------+----------------+----------------+----------------+
+ * | reserved | 8
+ * +----------------+----------------+----------------+----------------+
+ *
+ * id - uint16_t, event id, may be used by front
+ * type - uint8_t, type of the event
+ *
+ *
+ * Page flip complete event - event from back to front on page flip completed:
+ * 0 1 2 3 octet
+ * +----------------+----------------+----------------+----------------+
+ * | id | _EVT_PG_FLIP | reserved | 4
+ * +----------------+----------------+----------------+----------------+
+ * | reserved | 8
+ * +----------------+----------------+----------------+----------------+
+ * | fb_cookie low 32-bit | 12
+ * +----------------+----------------+----------------+----------------+
+ * | fb_cookie high 32-bit | 16
+ * +----------------+----------------+----------------+----------------+
+ * | reserved | 20
+ * +----------------+----------------+----------------+----------------+
+ * |/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/|
+ * +----------------+----------------+----------------+----------------+
+ * | reserved | 64
+ * +----------------+----------------+----------------+----------------+
+ */
+
+struct xendispl_pg_flip_evt {
+ uint64_t fb_cookie;
+};
+
+struct xendispl_req {
+ uint16_t id;
+ uint8_t operation;
+ uint8_t reserved[5];
+ union {
+ struct xendispl_dbuf_create_req dbuf_create;
+ struct xendispl_dbuf_destroy_req dbuf_destroy;
+ struct xendispl_fb_attach_req fb_attach;
+ struct xendispl_fb_detach_req fb_detach;
+ struct xendispl_set_config_req set_config;
+ struct xendispl_page_flip_req pg_flip;
+ uint8_t reserved[56];
+ } op;
+};
+
+struct xendispl_resp {
+ uint16_t id;
+ uint8_t operation;
+ uint8_t reserved;
+ int32_t status;
+ uint8_t reserved1[56];
+};
+
+struct xendispl_evt {
+ uint16_t id;
+ uint8_t type;
+ uint8_t reserved[5];
+ union {
+ struct xendispl_pg_flip_evt pg_flip;
+ uint8_t reserved[56];
+ } op;
+};
+
+DEFINE_RING_TYPES(xen_displif, struct xendispl_req, struct xendispl_resp);
+
+/*
+ ******************************************************************************
+ * Back to front events delivery
+ ******************************************************************************
+ * In order to deliver asynchronous events from back to front a shared page is
+ * allocated by front and its granted reference propagated to back via
+ * XenStore entries (evt-ring-ref/evt-event-channel).
+ * This page has a common header used by both front and back to synchronize
+ * access and control event's ring buffer, while back being a producer of the
+ * events and front being a consumer. The rest of the page after the header
+ * is used for event packets.
+ *
+ * Upon reception of an event(s) front may confirm its reception
+ * for either each event, group of events or none.
+ */
+
+struct xendispl_event_page {
+ uint32_t in_cons;
+ uint32_t in_prod;
+ uint8_t reserved[56];
+};
+
+#define XENDISPL_EVENT_PAGE_SIZE 4096
+#define XENDISPL_IN_RING_OFFS (sizeof(struct xendispl_event_page))
+#define XENDISPL_IN_RING_SIZE (XENDISPL_EVENT_PAGE_SIZE - XENDISPL_IN_RING_OFFS)
+#define XENDISPL_IN_RING_LEN (XENDISPL_IN_RING_SIZE / sizeof(struct xendispl_evt))
+#define XENDISPL_IN_RING(page) \
+ ((struct xendispl_evt *)((char *)(page) + XENDISPL_IN_RING_OFFS))
+#define XENDISPL_IN_RING_REF(page, idx) \
+ (XENDISPL_IN_RING((page))[(idx) % XENDISPL_IN_RING_LEN])
+
+#endif /* __XEN_PUBLIC_IO_DISPLIF_H__ */
+
+/*
+ * Local variables:
+ * mode: C
+ * c-file-style: "BSD"
+ * c-basic-offset: 4
+ * tab-width: 4
+ * indent-tabs-mode: nil
+ * End:
+ */
--
2.7.4
^ permalink raw reply related [flat|nested] 41+ messages in thread
* [For Linux 4/4] xen/displif: add ABI for para-virtual display
2017-04-07 8:30 [For Linux 0/4] PV protocol headers for Linux Kernel Oleksandr Andrushchenko
` (6 preceding siblings ...)
2017-04-07 8:30 ` [For Linux 4/4] xen/displif: add ABI for para-virtual display Oleksandr Andrushchenko
@ 2017-04-07 8:30 ` Oleksandr Andrushchenko
2017-04-07 8:47 ` [For Linux 0/4] PV protocol headers for Linux Kernel Oleksandr Andrushchenko
2017-04-07 8:47 ` Oleksandr Andrushchenko
9 siblings, 0 replies; 41+ messages in thread
From: Oleksandr Andrushchenko @ 2017-04-07 8:30 UTC (permalink / raw)
To: xen-devel
Cc: lars.kurth, sstabellini, vlad.babchuk, linux-kernel,
Oleksandr Andrushchenko, julien.grall, andrii.anisov, olekstysh,
andr2000, al1img, Oleksandr Grytsov, joculator
From: Oleksandr Andrushchenko <oleksandr_andrushchenko@epam.com>
This is the ABI for the two halves of a para-virtualized
display driver.
This protocol aims to provide a unified protocol which fits more
sophisticated use-cases than a framebuffer device can handle. At the
moment basic functionality is supported with the intention to extend:
o multiple dynamically allocated/destroyed framebuffers
o buffers of arbitrary sizes
o better configuration options including multiple display support
Note: existing fbif can be used together with displif running at the
same time, e.g. on Linux one provides framebuffer and another DRM/KMS
Future extensions to the existing protocol may include:
o allow display/connector cloning
o allow allocating objects other than display buffers
o add planes/overlays support
o support scaling
o support rotation
Note, that this protocol doesn't use ring macros for
bi-directional exchange (PV calls/9pfs) bacause:
o it statically defines the use of a single page
for the ring buffer
o it uses direct memory access to ring's contents
w/o memory copying
o re-uses the same idea that kbdif/fbif use
which for this use-case seems to be appropriate
==================================================
Rationale for introducing this protocol instead of
using the existing fbif:
==================================================
1. In/out event sizes
o fbif - 40 octets
o displif - 40 octets
This is only the initial version of the displif protocol
which means that there could be requests which will not fit
(WRT introducing some GPU related functionality
later on). In that case we cannot alter fbif sizes as we need to
be backward compatible an will be forced to handle those
apart of fbif.
2. Shared page
Displif doesn't use anything like struct xenfb_page, but
DEFINE_RING_TYPES(xen_displif, struct xendispl_req, struct
xendispl_resp) which is a better and more common way.
Output events use a shared page which only has in_cons and in_prod
and all the rest is used for incoming events. Here struct xenfb_page
could probably be used as is despite the fact that it only has a half
of a page for incoming events which is only 50 events. (consider
something like 60Hz display)
3. Amount of changes.
fbif only provides XENFB_TYPE_UPDATE and XENFB_TYPE_RESIZE
events, so it looks like it is easier to get fb support into displif
than vice versa. displif at the moment has 6 requests and 1 event,
multiple connector support, etc.
Reviewed-by: Konrad Rzeszutek Wilk <konrad.wilk@oracle.com>
Signed-off-by: Oleksandr Grytsov <oleksandr_grytsov@epam.com>
Signed-off-by: Oleksandr Andrushchenko <oleksandr_andrushchenko@epam.com>
---
include/xen/interface/io/displif.h | 864 +++++++++++++++++++++++++++++++++++++
1 file changed, 864 insertions(+)
create mode 100644 include/xen/interface/io/displif.h
diff --git a/include/xen/interface/io/displif.h b/include/xen/interface/io/displif.h
new file mode 100644
index 000000000000..ac2b489ecf27
--- /dev/null
+++ b/include/xen/interface/io/displif.h
@@ -0,0 +1,864 @@
+/******************************************************************************
+ * displif.h
+ *
+ * Unified display device I/O interface for Xen guest OSes.
+ *
+ * Permission is hereby granted, free of charge, to any person obtaining a copy
+ * of this software and associated documentation files (the "Software"), to
+ * deal in the Software without restriction, including without limitation the
+ * rights to use, copy, modify, merge, publish, distribute, sublicense, and/or
+ * sell copies of the Software, and to permit persons to whom the Software is
+ * furnished to do so, subject to the following conditions:
+ *
+ * The above copyright notice and this permission notice shall be included in
+ * all copies or substantial portions of the Software.
+ *
+ * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
+ * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
+ * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE
+ * AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER
+ * LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING
+ * FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER
+ * DEALINGS IN THE SOFTWARE.
+ *
+ * Copyright (C) 2016-2017 EPAM Systems Inc.
+ *
+ * Authors: Oleksandr Andrushchenko <oleksandr_andrushchenko@epam.com>
+ * Oleksandr Grytsov <oleksandr_grytsov@epam.com>
+ */
+
+#ifndef __XEN_PUBLIC_IO_DISPLIF_H__
+#define __XEN_PUBLIC_IO_DISPLIF_H__
+
+#include "ring.h"
+#include "../grant_table.h"
+
+/*
+ ******************************************************************************
+ * Protocol version
+ ******************************************************************************
+ */
+#define XENDISPL_PROTOCOL_VERSION "1"
+
+/*
+ ******************************************************************************
+ * Main features provided by the protocol
+ ******************************************************************************
+ * This protocol aims to provide a unified protocol which fits more
+ * sophisticated use-cases than a framebuffer device can handle. At the
+ * moment basic functionality is supported with the intention to be extended:
+ * o multiple dynamically allocated/destroyed framebuffers
+ * o buffers of arbitrary sizes
+ * o buffer allocation at either back or front end
+ * o better configuration options including multiple display support
+ *
+ * Note: existing fbif can be used together with displif running at the
+ * same time, e.g. on Linux one provides framebuffer and another DRM/KMS
+ *
+ * Note: display resolution (XenStore's "resolution" property) defines
+ * visible area of the virtual display. At the same time resolution of
+ * the display and frame buffers may differ: buffers can be smaller, equal
+ * or bigger than the visible area. This is to enable use-cases, where backend
+ * may do some post-processing of the display and frame buffers supplied,
+ * e.g. those buffers can be just a part of the final composition.
+ *
+ ******************************************************************************
+ * Direction of improvements
+ ******************************************************************************
+ * Future extensions to the existing protocol may include:
+ * o display/connector cloning
+ * o allocation of objects other than display buffers
+ * o plane/overlay support
+ * o scaling support
+ * o rotation support
+ *
+ ******************************************************************************
+ * Feature and Parameter Negotiation
+ ******************************************************************************
+ *
+ * Front->back notifications: when enqueuing a new request, sending a
+ * notification can be made conditional on xendispl_req (i.e., the generic
+ * hold-off mechanism provided by the ring macros). Backends must set
+ * xendispl_req appropriately (e.g., using RING_FINAL_CHECK_FOR_REQUESTS()).
+ *
+ * Back->front notifications: when enqueuing a new response, sending a
+ * notification can be made conditional on xendispl_resp (i.e., the generic
+ * hold-off mechanism provided by the ring macros). Frontends must set
+ * xendispl_resp appropriately (e.g., using RING_FINAL_CHECK_FOR_RESPONSES()).
+ *
+ * The two halves of a para-virtual display driver utilize nodes within
+ * XenStore to communicate capabilities and to negotiate operating parameters.
+ * This section enumerates these nodes which reside in the respective front and
+ * backend portions of XenStore, following the XenBus convention.
+ *
+ * All data in XenStore is stored as strings. Nodes specifying numeric
+ * values are encoded in decimal. Integer value ranges listed below are
+ * expressed as fixed sized integer types capable of storing the conversion
+ * of a properly formated node string, without loss of information.
+ *
+ ******************************************************************************
+ * Example configuration
+ ******************************************************************************
+ *
+ * Note: depending on the use-case backend can expose more display connectors
+ * than the underlying HW physically has by employing SW graphics compositors
+ *
+ * This is an example of backend and frontend configuration:
+ *
+ *--------------------------------- Backend -----------------------------------
+ *
+ * /local/domain/0/backend/vdispl/1/0/frontend-id = "1"
+ * /local/domain/0/backend/vdispl/1/0/frontend = "/local/domain/1/device/vdispl/0"
+ * /local/domain/0/backend/vdispl/1/0/state = "4"
+ * /local/domain/0/backend/vdispl/1/0/versions = "1,2"
+ *
+ *--------------------------------- Frontend ----------------------------------
+ *
+ * /local/domain/1/device/vdispl/0/backend-id = "0"
+ * /local/domain/1/device/vdispl/0/backend = "/local/domain/0/backend/vdispl/1/0"
+ * /local/domain/1/device/vdispl/0/state = "4"
+ * /local/domain/1/device/vdispl/0/version = "1"
+ * /local/domain/1/device/vdispl/0/be-alloc = "1"
+ *
+ *-------------------------- Connector 0 configuration ------------------------
+ *
+ * /local/domain/1/device/vdispl/0/0/resolution = "1920x1080"
+ * /local/domain/1/device/vdispl/0/0/req-ring-ref = "2832"
+ * /local/domain/1/device/vdispl/0/0/req-event-channel = "15"
+ * /local/domain/1/device/vdispl/0/0/evt-ring-ref = "387"
+ * /local/domain/1/device/vdispl/0/0/evt-event-channel = "16"
+ *
+ *-------------------------- Connector 1 configuration ------------------------
+ *
+ * /local/domain/1/device/vdispl/0/1/resolution = "800x600"
+ * /local/domain/1/device/vdispl/0/1/req-ring-ref = "2833"
+ * /local/domain/1/device/vdispl/0/1/req-event-channel = "17"
+ * /local/domain/1/device/vdispl/0/1/evt-ring-ref = "388"
+ * /local/domain/1/device/vdispl/0/1/evt-event-channel = "18"
+ *
+ ******************************************************************************
+ * Backend XenBus Nodes
+ ******************************************************************************
+ *
+ *----------------------------- Protocol version ------------------------------
+ *
+ * versions
+ * Values: <string>
+ *
+ * List of XENDISPL_LIST_SEPARATOR separated protocol versions supported
+ * by the backend. For example "1,2,3".
+ *
+ ******************************************************************************
+ * Frontend XenBus Nodes
+ ******************************************************************************
+ *
+ *-------------------------------- Addressing ---------------------------------
+ *
+ * dom-id
+ * Values: <uint16_t>
+ *
+ * Domain identifier.
+ *
+ * dev-id
+ * Values: <uint16_t>
+ *
+ * Device identifier.
+ *
+ * conn-idx
+ * Values: <uint8_t>
+ *
+ * Zero based contigous index of the connector.
+ * /local/domain/<dom-id>/device/vdispl/<dev-id>/<conn-idx>/...
+ *
+ *----------------------------- Protocol version ------------------------------
+ *
+ * version
+ * Values: <string>
+ *
+ * Protocol version, chosen among the ones supported by the backend.
+ *
+ *------------------------- Backend buffer allocation -------------------------
+ *
+ * be-alloc
+ * Values: "0", "1"
+ *
+ * If value is set to "1", then backend can be a buffer provider/allocator
+ * for this domain during XENDISPL_OP_DBUF_CREATE operation (see below
+ * for negotiation).
+ * If value is not "1" or omitted frontend must allocate buffers itself.
+ *
+ *----------------------------- Connector settings ----------------------------
+ *
+ * resolution
+ * Values: <width, uint32_t>x<height, uint32_t>
+ *
+ * Width and height of the connector in pixels separated by
+ * XENDISPL_RESOLUTION_SEPARATOR. This defines visible area of the
+ * display.
+ *
+ *------------------ Connector Request Transport Parameters -------------------
+ *
+ * This communication path is used to deliver requests from frontend to backend
+ * and get the corresponding responses from backend to frontend,
+ * set up per connector.
+ *
+ * req-event-channel
+ * Values: <uint32_t>
+ *
+ * The identifier of the Xen connector's control event channel
+ * used to signal activity in the ring buffer.
+ *
+ * req-ring-ref
+ * Values: <uint32_t>
+ *
+ * The Xen grant reference granting permission for the backend to map
+ * a sole page of connector's control ring buffer.
+ *
+ *------------------- Connector Event Transport Parameters --------------------
+ *
+ * This communication path is used to deliver asynchronous events from backend
+ * to frontend, set up per connector.
+ *
+ * evt-event-channel
+ * Values: <uint32_t>
+ *
+ * The identifier of the Xen connector's event channel
+ * used to signal activity in the ring buffer.
+ *
+ * evt-ring-ref
+ * Values: <uint32_t>
+ *
+ * The Xen grant reference granting permission for the backend to map
+ * a sole page of connector's event ring buffer.
+ */
+
+/*
+ ******************************************************************************
+ * STATE DIAGRAMS
+ ******************************************************************************
+ *
+ * Tool stack creates front and back state nodes with initial state
+ * XenbusStateInitialising.
+ * Tool stack creates and sets up frontend display configuration
+ * nodes per domain.
+ *
+ *-------------------------------- Normal flow --------------------------------
+ *
+ * Front Back
+ * ================================= =====================================
+ * XenbusStateInitialising XenbusStateInitialising
+ * o Query backend device identification
+ * data.
+ * o Open and validate backend device.
+ * |
+ * |
+ * V
+ * XenbusStateInitWait
+ *
+ * o Query frontend configuration
+ * o Allocate and initialize
+ * event channels per configured
+ * connector.
+ * o Publish transport parameters
+ * that will be in effect during
+ * this connection.
+ * |
+ * |
+ * V
+ * XenbusStateInitialised
+ *
+ * o Query frontend transport parameters.
+ * o Connect to the event channels.
+ * |
+ * |
+ * V
+ * XenbusStateConnected
+ *
+ * o Create and initialize OS
+ * virtual display connectors
+ * as per configuration.
+ * |
+ * |
+ * V
+ * XenbusStateConnected
+ *
+ * XenbusStateUnknown
+ * XenbusStateClosed
+ * XenbusStateClosing
+ * o Remove virtual display device
+ * o Remove event channels
+ * |
+ * |
+ * V
+ * XenbusStateClosed
+ *
+ *------------------------------- Recovery flow -------------------------------
+ *
+ * In case of frontend unrecoverable errors backend handles that as
+ * if frontend goes into the XenbusStateClosed state.
+ *
+ * In case of backend unrecoverable errors frontend tries removing
+ * the virtualized device. If this is possible at the moment of error,
+ * then frontend goes into the XenbusStateInitialising state and is ready for
+ * new connection with backend. If the virtualized device is still in use and
+ * cannot be removed, then frontend goes into the XenbusStateReconfiguring state
+ * until either the virtualized device is removed or backend initiates a new
+ * connection. On the virtualized device removal frontend goes into the
+ * XenbusStateInitialising state.
+ *
+ * Note on XenbusStateReconfiguring state of the frontend: if backend has
+ * unrecoverable errors then frontend cannot send requests to the backend
+ * and thus cannot provide functionality of the virtualized device anymore.
+ * After backend is back to normal the virtualized device may still hold some
+ * state: configuration in use, allocated buffers, client application state etc.
+ * In most cases, this will require frontend to implement complex recovery
+ * reconnect logic. Instead, by going into XenbusStateReconfiguring state,
+ * frontend will make sure no new clients of the virtualized device are
+ * accepted, allow existing client(s) to exit gracefully by signaling error
+ * state etc.
+ * Once all the clients are gone frontend can reinitialize the virtualized
+ * device and get into XenbusStateInitialising state again signaling the
+ * backend that a new connection can be made.
+ *
+ * There are multiple conditions possible under which frontend will go from
+ * XenbusStateReconfiguring into XenbusStateInitialising, some of them are OS
+ * specific. For example:
+ * 1. The underlying OS framework may provide callbacks to signal that the last
+ * client of the virtualized device has gone and the device can be removed
+ * 2. Frontend can schedule a deferred work (timer/tasklet/workqueue)
+ * to periodically check if this is the right time to re-try removal of
+ * the virtualized device.
+ * 3. By any other means.
+ *
+ ******************************************************************************
+ * REQUEST CODES
+ ******************************************************************************
+ * Request codes [0; 15] are reserved and must not be used
+ */
+
+#define XENDISPL_OP_DBUF_CREATE 0x10
+#define XENDISPL_OP_DBUF_DESTROY 0x11
+#define XENDISPL_OP_FB_ATTACH 0x12
+#define XENDISPL_OP_FB_DETACH 0x13
+#define XENDISPL_OP_SET_CONFIG 0x14
+#define XENDISPL_OP_PG_FLIP 0x15
+
+/*
+ ******************************************************************************
+ * EVENT CODES
+ ******************************************************************************
+ */
+#define XENDISPL_EVT_PG_FLIP 0x00
+
+/*
+ ******************************************************************************
+ * XENSTORE FIELD AND PATH NAME STRINGS, HELPERS
+ ******************************************************************************
+ */
+#define XENDISPL_DRIVER_NAME "vdispl"
+
+#define XENDISPL_LIST_SEPARATOR ","
+#define XENDISPL_RESOLUTION_SEPARATOR "x"
+
+#define XENDISPL_FIELD_BE_VERSIONS "versions"
+#define XENDISPL_FIELD_FE_VERSION "version"
+#define XENDISPL_FIELD_REQ_RING_REF "req-ring-ref"
+#define XENDISPL_FIELD_REQ_CHANNEL "req-event-channel"
+#define XENDISPL_FIELD_EVT_RING_REF "evt-ring-ref"
+#define XENDISPL_FIELD_EVT_CHANNEL "evt-event-channel"
+#define XENDISPL_FIELD_RESOLUTION "resolution"
+#define XENDISPL_FIELD_BE_ALLOC "be-alloc"
+
+/*
+ ******************************************************************************
+ * STATUS RETURN CODES
+ ******************************************************************************
+ *
+ * Status return code is zero on success and -XEN_EXX on failure.
+ *
+ ******************************************************************************
+ * Assumptions
+ ******************************************************************************
+ * o usage of grant reference 0 as invalid grant reference:
+ * grant reference 0 is valid, but never exposed to a PV driver,
+ * because of the fact it is already in use/reserved by the PV console.
+ * o all references in this document to page sizes must be treated
+ * as pages of size XEN_PAGE_SIZE unless otherwise noted.
+ *
+ ******************************************************************************
+ * Description of the protocol between frontend and backend driver
+ ******************************************************************************
+ *
+ * The two halves of a Para-virtual display driver communicate with
+ * each other using shared pages and event channels.
+ * Shared page contains a ring with request/response packets.
+ *
+ * All reserved fields in the structures below must be 0.
+ * Display buffers's cookie of value 0 is treated as invalid.
+ * Framebuffer's cookie of value 0 is treated as invalid.
+ *
+ * For all request/response/event packets that use cookies:
+ * dbuf_cookie - uint64_t, unique to guest domain value used by the backend
+ * to map remote display buffer to its local one
+ * fb_cookie - uint64_t, unique to guest domain value used by the backend
+ * to map remote framebuffer to its local one
+ *
+ *---------------------------------- Requests ---------------------------------
+ *
+ * All requests/responses, which are not connector specific, must be sent over
+ * control ring of the connector which has the index value of 0:
+ * /local/domain/<dom-id>/device/vdispl/<dev-id>/0/req-ring-ref
+ *
+ * All request packets have the same length (64 octets)
+ * All request packets have common header:
+ * 0 1 2 3 octet
+ * +----------------+----------------+----------------+----------------+
+ * | id | operation | reserved | 4
+ * +----------------+----------------+----------------+----------------+
+ * | reserved | 8
+ * +----------------+----------------+----------------+----------------+
+ * id - uint16_t, private guest value, echoed in response
+ * operation - uint8_t, operation code, XENDISPL_OP_???
+ *
+ * Request dbuf creation - request creation of a display buffer.
+ * 0 1 2 3 octet
+ * +----------------+----------------+----------------+----------------+
+ * | id |_OP_DBUF_CREATE | reserved | 4
+ * +----------------+----------------+----------------+----------------+
+ * | reserved | 8
+ * +----------------+----------------+----------------+----------------+
+ * | dbuf_cookie low 32-bit | 12
+ * +----------------+----------------+----------------+----------------+
+ * | dbuf_cookie high 32-bit | 16
+ * +----------------+----------------+----------------+----------------+
+ * | width | 20
+ * +----------------+----------------+----------------+----------------+
+ * | height | 24
+ * +----------------+----------------+----------------+----------------+
+ * | bpp | 28
+ * +----------------+----------------+----------------+----------------+
+ * | buffer_sz | 32
+ * +----------------+----------------+----------------+----------------+
+ * | flags | 36
+ * +----------------+----------------+----------------+----------------+
+ * | gref_directory | 40
+ * +----------------+----------------+----------------+----------------+
+ * | reserved | 44
+ * +----------------+----------------+----------------+----------------+
+ * |/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/|
+ * +----------------+----------------+----------------+----------------+
+ * | reserved | 64
+ * +----------------+----------------+----------------+----------------+
+ *
+ * Must be sent over control ring of the connector which has the index
+ * value of 0:
+ * /local/domain/<dom-id>/device/vdispl/<dev-id>/0/req-ring-ref
+ * All unused bits in flags field must be set to 0.
+ *
+ * An attempt to create multiple display buffers with the same dbuf_cookie is
+ * an error. dbuf_cookie can be re-used after destroying the corresponding
+ * display buffer.
+ *
+ * Width and height of the display buffers can be smaller, equal or bigger
+ * than the connector's resolution. Depth/pixel format of the individual
+ * buffers can differ as well.
+ *
+ * width - uint32_t, width in pixels
+ * height - uint32_t, height in pixels
+ * bpp - uint32_t, bits per pixel
+ * buffer_sz - uint32_t, buffer size to be allocated, octets
+ * flags - uint32_t, flags of the operation
+ * o XENDISPL_DBUF_FLG_REQ_ALLOC - if set, then backend is requested
+ * to allocate the buffer with the parameters provided in this request.
+ * Page directory is handled as follows:
+ * Frontend on request:
+ * o allocates pages for the directory (gref_directory,
+ * gref_dir_next_page(s)
+ * o grants permissions for the pages of the directory to the backend
+ * o sets gref_dir_next_page fields
+ * Backend on response:
+ * o grants permissions for the pages of the buffer allocated to
+ * the frontend
+ * o fills in page directory with grant references
+ * (gref[] in struct xendispl_page_directory)
+ * gref_directory - grant_ref_t, a reference to the first shared page
+ * describing shared buffer references. At least one page exists. If shared
+ * buffer size (buffer_sz) exceeds what can be addressed by this single page,
+ * then reference to the next page must be supplied (see gref_dir_next_page
+ * below)
+ */
+
+#define XENDISPL_DBUF_FLG_REQ_ALLOC (1 << 0)
+
+struct xendispl_dbuf_create_req {
+ uint64_t dbuf_cookie;
+ uint32_t width;
+ uint32_t height;
+ uint32_t bpp;
+ uint32_t buffer_sz;
+ uint32_t flags;
+ grant_ref_t gref_directory;
+};
+
+/*
+ * Shared page for XENDISPL_OP_DBUF_CREATE buffer descriptor (gref_directory in
+ * the request) employs a list of pages, describing all pages of the shared
+ * data buffer:
+ * 0 1 2 3 octet
+ * +----------------+----------------+----------------+----------------+
+ * | gref_dir_next_page | 4
+ * +----------------+----------------+----------------+----------------+
+ * | gref[0] | 8
+ * +----------------+----------------+----------------+----------------+
+ * |/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/|
+ * +----------------+----------------+----------------+----------------+
+ * | gref[i] | i*4+8
+ * +----------------+----------------+----------------+----------------+
+ * |/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/|
+ * +----------------+----------------+----------------+----------------+
+ * | gref[N - 1] | N*4+8
+ * +----------------+----------------+----------------+----------------+
+ *
+ * gref_dir_next_page - grant_ref_t, reference to the next page describing
+ * page directory. Must be 0 if there are no more pages in the list.
+ * gref[i] - grant_ref_t, reference to a shared page of the buffer
+ * allocated at XENDISPL_OP_DBUF_CREATE
+ *
+ * Number of grant_ref_t entries in the whole page directory is not
+ * passed, but instead can be calculated as:
+ * num_grefs_total = (XENDISPL_OP_DBUF_CREATE.buffer_sz + XEN_PAGE_SIZE - 1) /
+ * XEN_PAGE_SIZE
+ */
+
+struct xendispl_page_directory {
+ grant_ref_t gref_dir_next_page;
+ grant_ref_t gref[1]; /* Variable length */
+};
+
+/*
+ * Request dbuf destruction - destroy a previously allocated display buffer:
+ * 0 1 2 3 octet
+ * +----------------+----------------+----------------+----------------+
+ * | id |_OP_DBUF_DESTROY| reserved | 4
+ * +----------------+----------------+----------------+----------------+
+ * | reserved | 8
+ * +----------------+----------------+----------------+----------------+
+ * | dbuf_cookie low 32-bit | 12
+ * +----------------+----------------+----------------+----------------+
+ * | dbuf_cookie high 32-bit | 16
+ * +----------------+----------------+----------------+----------------+
+ * | reserved | 20
+ * +----------------+----------------+----------------+----------------+
+ * |/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/|
+ * +----------------+----------------+----------------+----------------+
+ * | reserved | 64
+ * +----------------+----------------+----------------+----------------+
+ *
+ * Must be sent over control ring of the connector which has the index
+ * value of 0:
+ * /local/domain/<dom-id>/device/vdispl/<dev-id>/0/req-ring-ref
+ */
+
+struct xendispl_dbuf_destroy_req {
+ uint64_t dbuf_cookie;
+};
+
+/*
+ * Request framebuffer attachment - request attachment of a framebuffer to
+ * previously created display buffer.
+ * 0 1 2 3 octet
+ * +----------------+----------------+----------------+----------------+
+ * | id | _OP_FB_ATTACH | reserved | 4
+ * +----------------+----------------+----------------+----------------+
+ * | reserved | 8
+ * +----------------+----------------+----------------+----------------+
+ * | dbuf_cookie low 32-bit | 12
+ * +----------------+----------------+----------------+----------------+
+ * | dbuf_cookie high 32-bit | 16
+ * +----------------+----------------+----------------+----------------+
+ * | fb_cookie low 32-bit | 20
+ * +----------------+----------------+----------------+----------------+
+ * | fb_cookie high 32-bit | 24
+ * +----------------+----------------+----------------+----------------+
+ * | width | 28
+ * +----------------+----------------+----------------+----------------+
+ * | height | 32
+ * +----------------+----------------+----------------+----------------+
+ * | pixel_format | 36
+ * +----------------+----------------+----------------+----------------+
+ * | reserved | 40
+ * +----------------+----------------+----------------+----------------+
+ * |/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/|
+ * +----------------+----------------+----------------+----------------+
+ * | reserved | 64
+ * +----------------+----------------+----------------+----------------+
+ *
+ * Must be sent over control ring of the connector which has the index
+ * value of 0:
+ * /local/domain/<dom-id>/device/vdispl/<dev-id>/0/req-ring-ref
+ * Width and height can be smaller, equal or bigger than the connector's
+ * resolution.
+ *
+ * An attempt to create multiple frame buffers with the same fb_cookie is
+ * an error. fb_cookie can be re-used after destroying the corresponding
+ * frame buffer.
+ *
+ * width - uint32_t, width in pixels
+ * height - uint32_t, height in pixels
+ * pixel_format - uint32_t, pixel format of the framebuffer, FOURCC code
+ */
+
+struct xendispl_fb_attach_req {
+ uint64_t dbuf_cookie;
+ uint64_t fb_cookie;
+ uint32_t width;
+ uint32_t height;
+ uint32_t pixel_format;
+};
+
+/*
+ * Request framebuffer detach - detach a previously
+ * attached framebuffer from the display buffer in request:
+ * 0 1 2 3 octet
+ * +----------------+----------------+----------------+----------------+
+ * | id | _OP_FB_DETACH | reserved | 4
+ * +----------------+----------------+----------------+----------------+
+ * | reserved | 8
+ * +----------------+----------------+----------------+----------------+
+ * | fb_cookie low 32-bit | 12
+ * +----------------+----------------+----------------+----------------+
+ * | fb_cookie high 32-bit | 16
+ * +----------------+----------------+----------------+----------------+
+ * | reserved | 20
+ * +----------------+----------------+----------------+----------------+
+ * |/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/|
+ * +----------------+----------------+----------------+----------------+
+ * | reserved | 64
+ * +----------------+----------------+----------------+----------------+
+ *
+ * Must be sent over control ring of the connector which has the index
+ * value of 0:
+ * /local/domain/<dom-id>/device/vdispl/<dev-id>/0/req-ring-ref
+ */
+
+struct xendispl_fb_detach_req {
+ uint64_t fb_cookie;
+};
+
+/*
+ * Request configuration set/reset - request to set or reset
+ * the configuration/mode of the display:
+ * 0 1 2 3 octet
+ * +----------------+----------------+----------------+----------------+
+ * | id | _OP_SET_CONFIG | reserved | 4
+ * +----------------+----------------+----------------+----------------+
+ * | reserved | 8
+ * +----------------+----------------+----------------+----------------+
+ * | fb_cookie low 32-bit | 12
+ * +----------------+----------------+----------------+----------------+
+ * | fb_cookie high 32-bit | 16
+ * +----------------+----------------+----------------+----------------+
+ * | x | 20
+ * +----------------+----------------+----------------+----------------+
+ * | y | 24
+ * +----------------+----------------+----------------+----------------+
+ * | width | 28
+ * +----------------+----------------+----------------+----------------+
+ * | height | 32
+ * +----------------+----------------+----------------+----------------+
+ * | bpp | 40
+ * +----------------+----------------+----------------+----------------+
+ * | reserved | 44
+ * +----------------+----------------+----------------+----------------+
+ * |/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/|
+ * +----------------+----------------+----------------+----------------+
+ * | reserved | 64
+ * +----------------+----------------+----------------+----------------+
+ *
+ * Pass all zeros to reset, otherwise command is treated as
+ * configuration set.
+ * Framebuffer's cookie defines which framebuffer/dbuf must be
+ * displayed while enabling display (applying configuration).
+ * x, y, width and height are bound by the connector's resolution and must not
+ * exceed it.
+ *
+ * x - uint32_t, starting position in pixels by X axis
+ * y - uint32_t, starting position in pixels by Y axis
+ * width - uint32_t, width in pixels
+ * height - uint32_t, height in pixels
+ * bpp - uint32_t, bits per pixel
+ */
+
+struct xendispl_set_config_req {
+ uint64_t fb_cookie;
+ uint32_t x;
+ uint32_t y;
+ uint32_t width;
+ uint32_t height;
+ uint32_t bpp;
+};
+
+/*
+ * Request page flip - request to flip a page identified by the framebuffer
+ * cookie:
+ * 0 1 2 3 octet
+ * +----------------+----------------+----------------+----------------+
+ * | id | _OP_PG_FLIP | reserved | 4
+ * +----------------+----------------+----------------+----------------+
+ * | reserved | 8
+ * +----------------+----------------+----------------+----------------+
+ * | fb_cookie low 32-bit | 12
+ * +----------------+----------------+----------------+----------------+
+ * | fb_cookie high 32-bit | 16
+ * +----------------+----------------+----------------+----------------+
+ * | reserved | 20
+ * +----------------+----------------+----------------+----------------+
+ * |/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/|
+ * +----------------+----------------+----------------+----------------+
+ * | reserved | 64
+ * +----------------+----------------+----------------+----------------+
+ */
+
+struct xendispl_page_flip_req {
+ uint64_t fb_cookie;
+};
+
+/*
+ *---------------------------------- Responses --------------------------------
+ *
+ * All response packets have the same length (64 octets)
+ *
+ * All response packets have common header:
+ * 0 1 2 3 octet
+ * +----------------+----------------+----------------+----------------+
+ * | id | reserved | 4
+ * +----------------+----------------+----------------+----------------+
+ * | status | 8
+ * +----------------+----------------+----------------+----------------+
+ * | reserved | 12
+ * +----------------+----------------+----------------+----------------+
+ * |/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/|
+ * +----------------+----------------+----------------+----------------+
+ * | reserved | 64
+ * +----------------+----------------+----------------+----------------+
+ *
+ * id - uint16_t, private guest value, echoed from request
+ * status - int32_t, response status, zero on success and -XEN_EXX on failure
+ *
+ *----------------------------------- Events ----------------------------------
+ *
+ * Events are sent via a shared page allocated by the front and propagated by
+ * evt-event-channel/evt-ring-ref XenStore entries
+ * All event packets have the same length (64 octets)
+ * All event packets have common header:
+ * 0 1 2 3 octet
+ * +----------------+----------------+----------------+----------------+
+ * | id | type | reserved | 4
+ * +----------------+----------------+----------------+----------------+
+ * | reserved | 8
+ * +----------------+----------------+----------------+----------------+
+ *
+ * id - uint16_t, event id, may be used by front
+ * type - uint8_t, type of the event
+ *
+ *
+ * Page flip complete event - event from back to front on page flip completed:
+ * 0 1 2 3 octet
+ * +----------------+----------------+----------------+----------------+
+ * | id | _EVT_PG_FLIP | reserved | 4
+ * +----------------+----------------+----------------+----------------+
+ * | reserved | 8
+ * +----------------+----------------+----------------+----------------+
+ * | fb_cookie low 32-bit | 12
+ * +----------------+----------------+----------------+----------------+
+ * | fb_cookie high 32-bit | 16
+ * +----------------+----------------+----------------+----------------+
+ * | reserved | 20
+ * +----------------+----------------+----------------+----------------+
+ * |/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/|
+ * +----------------+----------------+----------------+----------------+
+ * | reserved | 64
+ * +----------------+----------------+----------------+----------------+
+ */
+
+struct xendispl_pg_flip_evt {
+ uint64_t fb_cookie;
+};
+
+struct xendispl_req {
+ uint16_t id;
+ uint8_t operation;
+ uint8_t reserved[5];
+ union {
+ struct xendispl_dbuf_create_req dbuf_create;
+ struct xendispl_dbuf_destroy_req dbuf_destroy;
+ struct xendispl_fb_attach_req fb_attach;
+ struct xendispl_fb_detach_req fb_detach;
+ struct xendispl_set_config_req set_config;
+ struct xendispl_page_flip_req pg_flip;
+ uint8_t reserved[56];
+ } op;
+};
+
+struct xendispl_resp {
+ uint16_t id;
+ uint8_t operation;
+ uint8_t reserved;
+ int32_t status;
+ uint8_t reserved1[56];
+};
+
+struct xendispl_evt {
+ uint16_t id;
+ uint8_t type;
+ uint8_t reserved[5];
+ union {
+ struct xendispl_pg_flip_evt pg_flip;
+ uint8_t reserved[56];
+ } op;
+};
+
+DEFINE_RING_TYPES(xen_displif, struct xendispl_req, struct xendispl_resp);
+
+/*
+ ******************************************************************************
+ * Back to front events delivery
+ ******************************************************************************
+ * In order to deliver asynchronous events from back to front a shared page is
+ * allocated by front and its granted reference propagated to back via
+ * XenStore entries (evt-ring-ref/evt-event-channel).
+ * This page has a common header used by both front and back to synchronize
+ * access and control event's ring buffer, while back being a producer of the
+ * events and front being a consumer. The rest of the page after the header
+ * is used for event packets.
+ *
+ * Upon reception of an event(s) front may confirm its reception
+ * for either each event, group of events or none.
+ */
+
+struct xendispl_event_page {
+ uint32_t in_cons;
+ uint32_t in_prod;
+ uint8_t reserved[56];
+};
+
+#define XENDISPL_EVENT_PAGE_SIZE 4096
+#define XENDISPL_IN_RING_OFFS (sizeof(struct xendispl_event_page))
+#define XENDISPL_IN_RING_SIZE (XENDISPL_EVENT_PAGE_SIZE - XENDISPL_IN_RING_OFFS)
+#define XENDISPL_IN_RING_LEN (XENDISPL_IN_RING_SIZE / sizeof(struct xendispl_evt))
+#define XENDISPL_IN_RING(page) \
+ ((struct xendispl_evt *)((char *)(page) + XENDISPL_IN_RING_OFFS))
+#define XENDISPL_IN_RING_REF(page, idx) \
+ (XENDISPL_IN_RING((page))[(idx) % XENDISPL_IN_RING_LEN])
+
+#endif /* __XEN_PUBLIC_IO_DISPLIF_H__ */
+
+/*
+ * Local variables:
+ * mode: C
+ * c-file-style: "BSD"
+ * c-basic-offset: 4
+ * tab-width: 4
+ * indent-tabs-mode: nil
+ * End:
+ */
--
2.7.4
_______________________________________________
Xen-devel mailing list
Xen-devel@lists.xen.org
https://lists.xen.org/xen-devel
^ permalink raw reply related [flat|nested] 41+ messages in thread
* Re: [For Linux 0/4] PV protocol headers for Linux Kernel
2017-04-07 8:30 [For Linux 0/4] PV protocol headers for Linux Kernel Oleksandr Andrushchenko
` (7 preceding siblings ...)
2017-04-07 8:30 ` Oleksandr Andrushchenko
@ 2017-04-07 8:47 ` Oleksandr Andrushchenko
2017-04-07 8:47 ` Oleksandr Andrushchenko
9 siblings, 0 replies; 41+ messages in thread
From: Oleksandr Andrushchenko @ 2017-04-07 8:47 UTC (permalink / raw)
To: xen-devel
Cc: joculator, al1img, vlad.babchuk, andrii.anisov, olekstysh,
konrad.wilk, lars.kurth, julien.grall, sstabellini, linux-kernel,
Oleksandr Andrushchenko, Juergen Gross
+Juergen
On 04/07/2017 11:30 AM, Oleksandr Andrushchenko wrote:
> From: Oleksandr Andrushchenko <oleksandr_andrushchenko@epam.com>
>
> Hi, all!
>
> This patch series adds/updates para-virtual device
> protocols for Linux Kernel (headers):
> o kbdif (multitouch support added)
> o sndif - sound (new)
> o displif - display (new)
>
> No changes, but Linux Kernel coding style applied.
>
> Thank you,
> Oleksandr
>
> Oleksandr Andrushchenko (4):
> xen/kbdif: update protocol description
> xen/kbdif: add multi-touch support
> xen/sndif: add sound-device ABI
> xen/displif: add ABI for para-virtual display
>
> include/xen/interface/io/displif.h | 864 +++++++++++++++++++++++++++++++++++++
> include/xen/interface/io/kbdif.h | 468 ++++++++++++++++++--
> include/xen/interface/io/sndif.h | 803 ++++++++++++++++++++++++++++++++++
> 3 files changed, 2108 insertions(+), 27 deletions(-)
> create mode 100644 include/xen/interface/io/displif.h
> create mode 100644 include/xen/interface/io/sndif.h
>
^ permalink raw reply [flat|nested] 41+ messages in thread
* Re: [For Linux 0/4] PV protocol headers for Linux Kernel
2017-04-07 8:30 [For Linux 0/4] PV protocol headers for Linux Kernel Oleksandr Andrushchenko
` (8 preceding siblings ...)
2017-04-07 8:47 ` [For Linux 0/4] PV protocol headers for Linux Kernel Oleksandr Andrushchenko
@ 2017-04-07 8:47 ` Oleksandr Andrushchenko
9 siblings, 0 replies; 41+ messages in thread
From: Oleksandr Andrushchenko @ 2017-04-07 8:47 UTC (permalink / raw)
To: xen-devel
Cc: Juergen Gross, lars.kurth, sstabellini, vlad.babchuk,
linux-kernel, Oleksandr Andrushchenko, julien.grall,
andrii.anisov, olekstysh, al1img, joculator
+Juergen
On 04/07/2017 11:30 AM, Oleksandr Andrushchenko wrote:
> From: Oleksandr Andrushchenko <oleksandr_andrushchenko@epam.com>
>
> Hi, all!
>
> This patch series adds/updates para-virtual device
> protocols for Linux Kernel (headers):
> o kbdif (multitouch support added)
> o sndif - sound (new)
> o displif - display (new)
>
> No changes, but Linux Kernel coding style applied.
>
> Thank you,
> Oleksandr
>
> Oleksandr Andrushchenko (4):
> xen/kbdif: update protocol description
> xen/kbdif: add multi-touch support
> xen/sndif: add sound-device ABI
> xen/displif: add ABI for para-virtual display
>
> include/xen/interface/io/displif.h | 864 +++++++++++++++++++++++++++++++++++++
> include/xen/interface/io/kbdif.h | 468 ++++++++++++++++++--
> include/xen/interface/io/sndif.h | 803 ++++++++++++++++++++++++++++++++++
> 3 files changed, 2108 insertions(+), 27 deletions(-)
> create mode 100644 include/xen/interface/io/displif.h
> create mode 100644 include/xen/interface/io/sndif.h
>
_______________________________________________
Xen-devel mailing list
Xen-devel@lists.xen.org
https://lists.xen.org/xen-devel
^ permalink raw reply [flat|nested] 41+ messages in thread
* Re: [Xen-devel] [For Linux 1/4] xen/kbdif: update protocol description
2017-04-07 8:30 ` Oleksandr Andrushchenko
@ 2017-04-07 11:29 ` Juergen Gross
2017-04-07 12:45 ` Konrad Rzeszutek Wilk
2017-04-07 12:45 ` [Xen-devel] " Konrad Rzeszutek Wilk
2017-04-07 11:29 ` Juergen Gross
1 sibling, 2 replies; 41+ messages in thread
From: Juergen Gross @ 2017-04-07 11:29 UTC (permalink / raw)
To: Oleksandr Andrushchenko, xen-devel
Cc: lars.kurth, sstabellini, vlad.babchuk, linux-kernel,
Oleksandr Andrushchenko, julien.grall, andrii.anisov, olekstysh,
al1img, joculator, Boris Ostrovsky
On 07/04/17 10:30, Oleksandr Andrushchenko wrote:
> From: Oleksandr Andrushchenko <oleksandr_andrushchenko@epam.com>
>
> The patch clarifies the protocol that is used by the PV keyboard
> drivers.
>
> Reviewed-by: Konrad Rzeszutek Wilk <konrad.wilk@oracle.com>
> Reviewed-by: Stefano Stabellini <sstabellini@kernel.org>
> Signed-off-by: Oleksandr Andrushchenko <oleksandr_andrushchenko@epam.com>
Acked-by: Juergen Gross <jgross@suse.com>
I don't think its appropriate to keep the R-b:'s from the Xen
commits for the Linux kernel.
I can remove them when committing in case no V2 will be necessary.
Added Boris.
Juergen
^ permalink raw reply [flat|nested] 41+ messages in thread
* Re: [For Linux 1/4] xen/kbdif: update protocol description
2017-04-07 8:30 ` Oleksandr Andrushchenko
2017-04-07 11:29 ` [Xen-devel] " Juergen Gross
@ 2017-04-07 11:29 ` Juergen Gross
1 sibling, 0 replies; 41+ messages in thread
From: Juergen Gross @ 2017-04-07 11:29 UTC (permalink / raw)
To: Oleksandr Andrushchenko, xen-devel
Cc: lars.kurth, sstabellini, Oleksandr Andrushchenko, vlad.babchuk,
linux-kernel, andrii.anisov, olekstysh, julien.grall, al1img,
Boris Ostrovsky, joculator
On 07/04/17 10:30, Oleksandr Andrushchenko wrote:
> From: Oleksandr Andrushchenko <oleksandr_andrushchenko@epam.com>
>
> The patch clarifies the protocol that is used by the PV keyboard
> drivers.
>
> Reviewed-by: Konrad Rzeszutek Wilk <konrad.wilk@oracle.com>
> Reviewed-by: Stefano Stabellini <sstabellini@kernel.org>
> Signed-off-by: Oleksandr Andrushchenko <oleksandr_andrushchenko@epam.com>
Acked-by: Juergen Gross <jgross@suse.com>
I don't think its appropriate to keep the R-b:'s from the Xen
commits for the Linux kernel.
I can remove them when committing in case no V2 will be necessary.
Added Boris.
Juergen
_______________________________________________
Xen-devel mailing list
Xen-devel@lists.xen.org
https://lists.xen.org/xen-devel
^ permalink raw reply [flat|nested] 41+ messages in thread
* Re: [Xen-devel] [For Linux 2/4] xen/kbdif: add multi-touch support
2017-04-07 8:30 ` Oleksandr Andrushchenko
2017-04-07 11:30 ` Juergen Gross
@ 2017-04-07 11:30 ` Juergen Gross
1 sibling, 0 replies; 41+ messages in thread
From: Juergen Gross @ 2017-04-07 11:30 UTC (permalink / raw)
To: Oleksandr Andrushchenko, xen-devel
Cc: lars.kurth, sstabellini, vlad.babchuk, linux-kernel,
Oleksandr Andrushchenko, julien.grall, andrii.anisov, olekstysh,
al1img, joculator, Boris Ostrovsky
On 07/04/17 10:30, Oleksandr Andrushchenko wrote:
> From: Oleksandr Andrushchenko <oleksandr_andrushchenko@epam.com>
>
> Multi-touch fields re-use the page that is used by the other features
> which means that you can interleave multi-touch, motion, and key
> events.
>
> Reviewed-by: Konrad Rzeszutek Wilk <konrad.wilk@oracle.com>
> Reviewed-by: Stefano Stabellini <sstabellini@kernel.org>
> Signed-off-by: Oleksandr Andrushchenko <oleksandr_andrushchenko@epam.com>
Acked-by: Juergen Gross <jgross@suse.com>
Juergen
^ permalink raw reply [flat|nested] 41+ messages in thread
* Re: [For Linux 2/4] xen/kbdif: add multi-touch support
2017-04-07 8:30 ` Oleksandr Andrushchenko
@ 2017-04-07 11:30 ` Juergen Gross
2017-04-07 11:30 ` [Xen-devel] " Juergen Gross
1 sibling, 0 replies; 41+ messages in thread
From: Juergen Gross @ 2017-04-07 11:30 UTC (permalink / raw)
To: Oleksandr Andrushchenko, xen-devel
Cc: lars.kurth, sstabellini, Oleksandr Andrushchenko, vlad.babchuk,
linux-kernel, andrii.anisov, olekstysh, julien.grall, al1img,
Boris Ostrovsky, joculator
On 07/04/17 10:30, Oleksandr Andrushchenko wrote:
> From: Oleksandr Andrushchenko <oleksandr_andrushchenko@epam.com>
>
> Multi-touch fields re-use the page that is used by the other features
> which means that you can interleave multi-touch, motion, and key
> events.
>
> Reviewed-by: Konrad Rzeszutek Wilk <konrad.wilk@oracle.com>
> Reviewed-by: Stefano Stabellini <sstabellini@kernel.org>
> Signed-off-by: Oleksandr Andrushchenko <oleksandr_andrushchenko@epam.com>
Acked-by: Juergen Gross <jgross@suse.com>
Juergen
_______________________________________________
Xen-devel mailing list
Xen-devel@lists.xen.org
https://lists.xen.org/xen-devel
^ permalink raw reply [flat|nested] 41+ messages in thread
* Re: [Xen-devel] [For Linux 3/4] xen/sndif: add sound-device ABI
2017-04-07 8:30 ` Oleksandr Andrushchenko
2017-04-07 11:31 ` Juergen Gross
@ 2017-04-07 11:31 ` Juergen Gross
1 sibling, 0 replies; 41+ messages in thread
From: Juergen Gross @ 2017-04-07 11:31 UTC (permalink / raw)
To: Oleksandr Andrushchenko, xen-devel
Cc: lars.kurth, Iurii Konovalenko, sstabellini, vlad.babchuk,
linux-kernel, Oleksandr Dmytryshyn, Oleksandr Andrushchenko,
julien.grall, andrii.anisov, olekstysh, al1img,
Oleksandr Grytsov, joculator, Boris Ostrovsky
On 07/04/17 10:30, Oleksandr Andrushchenko wrote:
> From: Oleksandr Andrushchenko <oleksandr_andrushchenko@epam.com>
>
> Add ABI for the two halves of a para-virtualized
> sound driver to communicate with each other.
>
> The ABI allows implementing audio playback and capture as
> well as volume control and possibility to mute/unmute
> audio sources.
>
> Note: depending on the use-case backend can expose more sound
> cards and PCM devices/streams than the underlying HW physically
> has by employing SW mixers, configuring virtual sound streams,
> channels etc. Thus, allowing fine tunned configurations per
> frontend.
>
> Reviewed-by: Konrad Rzeszutek Wilk <konrad.wilk@oracle.com>
> Signed-off-by: Oleksandr Andrushchenko <oleksandr_andrushchenko@epam.com>
> Signed-off-by: Oleksandr Grytsov <oleksandr_grytsov@epam.com>
> Signed-off-by: Oleksandr Dmytryshyn <oleksandr.dmytryshyn@globallogic.com>
> Signed-off-by: Iurii Konovalenko <iurii.konovalenko@globallogic.com>
Acked-by: Juergen Gross <jgross@suse.com>
Juergen
^ permalink raw reply [flat|nested] 41+ messages in thread
* Re: [For Linux 3/4] xen/sndif: add sound-device ABI
2017-04-07 8:30 ` Oleksandr Andrushchenko
@ 2017-04-07 11:31 ` Juergen Gross
2017-04-07 11:31 ` [Xen-devel] " Juergen Gross
1 sibling, 0 replies; 41+ messages in thread
From: Juergen Gross @ 2017-04-07 11:31 UTC (permalink / raw)
To: Oleksandr Andrushchenko, xen-devel
Cc: lars.kurth, Iurii Konovalenko, sstabellini,
Oleksandr Andrushchenko, vlad.babchuk, Oleksandr Dmytryshyn,
linux-kernel, andrii.anisov, olekstysh, julien.grall,
Oleksandr Grytsov, al1img, Boris Ostrovsky, joculator
On 07/04/17 10:30, Oleksandr Andrushchenko wrote:
> From: Oleksandr Andrushchenko <oleksandr_andrushchenko@epam.com>
>
> Add ABI for the two halves of a para-virtualized
> sound driver to communicate with each other.
>
> The ABI allows implementing audio playback and capture as
> well as volume control and possibility to mute/unmute
> audio sources.
>
> Note: depending on the use-case backend can expose more sound
> cards and PCM devices/streams than the underlying HW physically
> has by employing SW mixers, configuring virtual sound streams,
> channels etc. Thus, allowing fine tunned configurations per
> frontend.
>
> Reviewed-by: Konrad Rzeszutek Wilk <konrad.wilk@oracle.com>
> Signed-off-by: Oleksandr Andrushchenko <oleksandr_andrushchenko@epam.com>
> Signed-off-by: Oleksandr Grytsov <oleksandr_grytsov@epam.com>
> Signed-off-by: Oleksandr Dmytryshyn <oleksandr.dmytryshyn@globallogic.com>
> Signed-off-by: Iurii Konovalenko <iurii.konovalenko@globallogic.com>
Acked-by: Juergen Gross <jgross@suse.com>
Juergen
_______________________________________________
Xen-devel mailing list
Xen-devel@lists.xen.org
https://lists.xen.org/xen-devel
^ permalink raw reply [flat|nested] 41+ messages in thread
* Re: [Xen-devel] [For Linux 4/4] xen/displif: add ABI for para-virtual display
2017-04-07 8:30 ` [For Linux 4/4] xen/displif: add ABI for para-virtual display Oleksandr Andrushchenko
@ 2017-04-07 11:31 ` Juergen Gross
2017-04-07 11:31 ` Juergen Gross
` (2 subsequent siblings)
3 siblings, 0 replies; 41+ messages in thread
From: Juergen Gross @ 2017-04-07 11:31 UTC (permalink / raw)
To: Oleksandr Andrushchenko, xen-devel
Cc: lars.kurth, sstabellini, vlad.babchuk, linux-kernel,
Oleksandr Andrushchenko, julien.grall, andrii.anisov, olekstysh,
al1img, Oleksandr Grytsov, joculator, Boris Ostrovsky
On 07/04/17 10:30, Oleksandr Andrushchenko wrote:
> From: Oleksandr Andrushchenko <oleksandr_andrushchenko@epam.com>
>
> This is the ABI for the two halves of a para-virtualized
> display driver.
>
> This protocol aims to provide a unified protocol which fits more
> sophisticated use-cases than a framebuffer device can handle. At the
> moment basic functionality is supported with the intention to extend:
> o multiple dynamically allocated/destroyed framebuffers
> o buffers of arbitrary sizes
> o better configuration options including multiple display support
>
> Note: existing fbif can be used together with displif running at the
> same time, e.g. on Linux one provides framebuffer and another DRM/KMS
>
> Future extensions to the existing protocol may include:
> o allow display/connector cloning
> o allow allocating objects other than display buffers
> o add planes/overlays support
> o support scaling
> o support rotation
>
> Note, that this protocol doesn't use ring macros for
> bi-directional exchange (PV calls/9pfs) bacause:
> o it statically defines the use of a single page
> for the ring buffer
> o it uses direct memory access to ring's contents
> w/o memory copying
> o re-uses the same idea that kbdif/fbif use
> which for this use-case seems to be appropriate
>
> ==================================================
> Rationale for introducing this protocol instead of
> using the existing fbif:
> ==================================================
>
> 1. In/out event sizes
> o fbif - 40 octets
> o displif - 40 octets
> This is only the initial version of the displif protocol
> which means that there could be requests which will not fit
> (WRT introducing some GPU related functionality
> later on). In that case we cannot alter fbif sizes as we need to
> be backward compatible an will be forced to handle those
> apart of fbif.
>
> 2. Shared page
> Displif doesn't use anything like struct xenfb_page, but
> DEFINE_RING_TYPES(xen_displif, struct xendispl_req, struct
> xendispl_resp) which is a better and more common way.
> Output events use a shared page which only has in_cons and in_prod
> and all the rest is used for incoming events. Here struct xenfb_page
> could probably be used as is despite the fact that it only has a half
> of a page for incoming events which is only 50 events. (consider
> something like 60Hz display)
>
> 3. Amount of changes.
> fbif only provides XENFB_TYPE_UPDATE and XENFB_TYPE_RESIZE
> events, so it looks like it is easier to get fb support into displif
> than vice versa. displif at the moment has 6 requests and 1 event,
> multiple connector support, etc.
>
> Reviewed-by: Konrad Rzeszutek Wilk <konrad.wilk@oracle.com>
> Signed-off-by: Oleksandr Grytsov <oleksandr_grytsov@epam.com>
> Signed-off-by: Oleksandr Andrushchenko <oleksandr_andrushchenko@epam.com>
Acked-by: Juergen Gross <jgross@suse.com>
Juergen
^ permalink raw reply [flat|nested] 41+ messages in thread
* Re: [For Linux 4/4] xen/displif: add ABI for para-virtual display
2017-04-07 8:30 ` [For Linux 4/4] xen/displif: add ABI for para-virtual display Oleksandr Andrushchenko
2017-04-07 11:31 ` [Xen-devel] " Juergen Gross
@ 2017-04-07 11:31 ` Juergen Gross
2017-04-07 13:50 ` Julien Grall
2017-04-07 13:50 ` Julien Grall
3 siblings, 0 replies; 41+ messages in thread
From: Juergen Gross @ 2017-04-07 11:31 UTC (permalink / raw)
To: Oleksandr Andrushchenko, xen-devel
Cc: lars.kurth, sstabellini, Oleksandr Andrushchenko, vlad.babchuk,
linux-kernel, andrii.anisov, olekstysh, julien.grall,
Oleksandr Grytsov, al1img, Boris Ostrovsky, joculator
On 07/04/17 10:30, Oleksandr Andrushchenko wrote:
> From: Oleksandr Andrushchenko <oleksandr_andrushchenko@epam.com>
>
> This is the ABI for the two halves of a para-virtualized
> display driver.
>
> This protocol aims to provide a unified protocol which fits more
> sophisticated use-cases than a framebuffer device can handle. At the
> moment basic functionality is supported with the intention to extend:
> o multiple dynamically allocated/destroyed framebuffers
> o buffers of arbitrary sizes
> o better configuration options including multiple display support
>
> Note: existing fbif can be used together with displif running at the
> same time, e.g. on Linux one provides framebuffer and another DRM/KMS
>
> Future extensions to the existing protocol may include:
> o allow display/connector cloning
> o allow allocating objects other than display buffers
> o add planes/overlays support
> o support scaling
> o support rotation
>
> Note, that this protocol doesn't use ring macros for
> bi-directional exchange (PV calls/9pfs) bacause:
> o it statically defines the use of a single page
> for the ring buffer
> o it uses direct memory access to ring's contents
> w/o memory copying
> o re-uses the same idea that kbdif/fbif use
> which for this use-case seems to be appropriate
>
> ==================================================
> Rationale for introducing this protocol instead of
> using the existing fbif:
> ==================================================
>
> 1. In/out event sizes
> o fbif - 40 octets
> o displif - 40 octets
> This is only the initial version of the displif protocol
> which means that there could be requests which will not fit
> (WRT introducing some GPU related functionality
> later on). In that case we cannot alter fbif sizes as we need to
> be backward compatible an will be forced to handle those
> apart of fbif.
>
> 2. Shared page
> Displif doesn't use anything like struct xenfb_page, but
> DEFINE_RING_TYPES(xen_displif, struct xendispl_req, struct
> xendispl_resp) which is a better and more common way.
> Output events use a shared page which only has in_cons and in_prod
> and all the rest is used for incoming events. Here struct xenfb_page
> could probably be used as is despite the fact that it only has a half
> of a page for incoming events which is only 50 events. (consider
> something like 60Hz display)
>
> 3. Amount of changes.
> fbif only provides XENFB_TYPE_UPDATE and XENFB_TYPE_RESIZE
> events, so it looks like it is easier to get fb support into displif
> than vice versa. displif at the moment has 6 requests and 1 event,
> multiple connector support, etc.
>
> Reviewed-by: Konrad Rzeszutek Wilk <konrad.wilk@oracle.com>
> Signed-off-by: Oleksandr Grytsov <oleksandr_grytsov@epam.com>
> Signed-off-by: Oleksandr Andrushchenko <oleksandr_andrushchenko@epam.com>
Acked-by: Juergen Gross <jgross@suse.com>
Juergen
_______________________________________________
Xen-devel mailing list
Xen-devel@lists.xen.org
https://lists.xen.org/xen-devel
^ permalink raw reply [flat|nested] 41+ messages in thread
* Re: [Xen-devel] [For Linux 1/4] xen/kbdif: update protocol description
2017-04-07 11:29 ` [Xen-devel] " Juergen Gross
2017-04-07 12:45 ` Konrad Rzeszutek Wilk
@ 2017-04-07 12:45 ` Konrad Rzeszutek Wilk
2017-04-07 12:58 ` Oleksandr Andrushchenko
2017-04-07 12:58 ` [Xen-devel] " Oleksandr Andrushchenko
1 sibling, 2 replies; 41+ messages in thread
From: Konrad Rzeszutek Wilk @ 2017-04-07 12:45 UTC (permalink / raw)
To: Juergen Gross
Cc: Oleksandr Andrushchenko, xen-devel, lars.kurth, sstabellini,
Oleksandr Andrushchenko, vlad.babchuk, linux-kernel,
andrii.anisov, olekstysh, julien.grall, al1img, Boris Ostrovsky,
joculator
On Fri, Apr 07, 2017 at 01:29:27PM +0200, Juergen Gross wrote:
> On 07/04/17 10:30, Oleksandr Andrushchenko wrote:
> > From: Oleksandr Andrushchenko <oleksandr_andrushchenko@epam.com>
> >
> > The patch clarifies the protocol that is used by the PV keyboard
> > drivers.
> >
> > Reviewed-by: Konrad Rzeszutek Wilk <konrad.wilk@oracle.com>
> > Reviewed-by: Stefano Stabellini <sstabellini@kernel.org>
> > Signed-off-by: Oleksandr Andrushchenko <oleksandr_andrushchenko@epam.com>
>
> Acked-by: Juergen Gross <jgross@suse.com>
>
> I don't think its appropriate to keep the R-b:'s from the Xen
> commits for the Linux kernel.
<nods>
>
> I can remove them when committing in case no V2 will be necessary.
Thank you!
Oleksandr, thank you for doing this!
>
> Added Boris.
>
>
> Juergen
>
> _______________________________________________
> Xen-devel mailing list
> Xen-devel@lists.xen.org
> https://lists.xen.org/xen-devel
^ permalink raw reply [flat|nested] 41+ messages in thread
* Re: [For Linux 1/4] xen/kbdif: update protocol description
2017-04-07 11:29 ` [Xen-devel] " Juergen Gross
@ 2017-04-07 12:45 ` Konrad Rzeszutek Wilk
2017-04-07 12:45 ` [Xen-devel] " Konrad Rzeszutek Wilk
1 sibling, 0 replies; 41+ messages in thread
From: Konrad Rzeszutek Wilk @ 2017-04-07 12:45 UTC (permalink / raw)
To: Juergen Gross
Cc: lars.kurth, sstabellini, Oleksandr Andrushchenko,
Oleksandr Andrushchenko, linux-kernel, andrii.anisov, olekstysh,
vlad.babchuk, julien.grall, al1img, xen-devel, Boris Ostrovsky,
joculator
On Fri, Apr 07, 2017 at 01:29:27PM +0200, Juergen Gross wrote:
> On 07/04/17 10:30, Oleksandr Andrushchenko wrote:
> > From: Oleksandr Andrushchenko <oleksandr_andrushchenko@epam.com>
> >
> > The patch clarifies the protocol that is used by the PV keyboard
> > drivers.
> >
> > Reviewed-by: Konrad Rzeszutek Wilk <konrad.wilk@oracle.com>
> > Reviewed-by: Stefano Stabellini <sstabellini@kernel.org>
> > Signed-off-by: Oleksandr Andrushchenko <oleksandr_andrushchenko@epam.com>
>
> Acked-by: Juergen Gross <jgross@suse.com>
>
> I don't think its appropriate to keep the R-b:'s from the Xen
> commits for the Linux kernel.
<nods>
>
> I can remove them when committing in case no V2 will be necessary.
Thank you!
Oleksandr, thank you for doing this!
>
> Added Boris.
>
>
> Juergen
>
> _______________________________________________
> Xen-devel mailing list
> Xen-devel@lists.xen.org
> https://lists.xen.org/xen-devel
_______________________________________________
Xen-devel mailing list
Xen-devel@lists.xen.org
https://lists.xen.org/xen-devel
^ permalink raw reply [flat|nested] 41+ messages in thread
* Re: [Xen-devel] [For Linux 1/4] xen/kbdif: update protocol description
2017-04-07 12:45 ` [Xen-devel] " Konrad Rzeszutek Wilk
2017-04-07 12:58 ` Oleksandr Andrushchenko
@ 2017-04-07 12:58 ` Oleksandr Andrushchenko
1 sibling, 0 replies; 41+ messages in thread
From: Oleksandr Andrushchenko @ 2017-04-07 12:58 UTC (permalink / raw)
To: Konrad Rzeszutek Wilk, Juergen Gross
Cc: xen-devel, lars.kurth, sstabellini, Oleksandr Andrushchenko,
vlad.babchuk, linux-kernel, andrii.anisov, olekstysh,
julien.grall, al1img, Boris Ostrovsky, joculator
On 04/07/2017 03:45 PM, Konrad Rzeszutek Wilk wrote:
> On Fri, Apr 07, 2017 at 01:29:27PM +0200, Juergen Gross wrote:
>> On 07/04/17 10:30, Oleksandr Andrushchenko wrote:
>>> From: Oleksandr Andrushchenko <oleksandr_andrushchenko@epam.com>
>>>
>>> The patch clarifies the protocol that is used by the PV keyboard
>>> drivers.
>>>
>>> Reviewed-by: Konrad Rzeszutek Wilk <konrad.wilk@oracle.com>
>>> Reviewed-by: Stefano Stabellini <sstabellini@kernel.org>
>>> Signed-off-by: Oleksandr Andrushchenko <oleksandr_andrushchenko@epam.com>
>> Acked-by: Juergen Gross <jgross@suse.com>
>>
>> I don't think its appropriate to keep the R-b:'s from the Xen
>> commits for the Linux kernel.
> <nods>
>
>> I can remove them when committing in case no V2 will be necessary.
Please, do me a favor
> Thank you!
>
> Oleksandr, thank you for doing this!
My pleasure
>> Added Boris.
>>
>>
>> Juergen
>>
>> _______________________________________________
>> Xen-devel mailing list
>> Xen-devel@lists.xen.org
>> https://lists.xen.org/xen-devel
Thank you,
Oleksandr
^ permalink raw reply [flat|nested] 41+ messages in thread
* Re: [For Linux 1/4] xen/kbdif: update protocol description
2017-04-07 12:45 ` [Xen-devel] " Konrad Rzeszutek Wilk
@ 2017-04-07 12:58 ` Oleksandr Andrushchenko
2017-04-07 12:58 ` [Xen-devel] " Oleksandr Andrushchenko
1 sibling, 0 replies; 41+ messages in thread
From: Oleksandr Andrushchenko @ 2017-04-07 12:58 UTC (permalink / raw)
To: Konrad Rzeszutek Wilk, Juergen Gross
Cc: lars.kurth, sstabellini, Oleksandr Andrushchenko, vlad.babchuk,
linux-kernel, andrii.anisov, olekstysh, julien.grall, al1img,
xen-devel, Boris Ostrovsky, joculator
On 04/07/2017 03:45 PM, Konrad Rzeszutek Wilk wrote:
> On Fri, Apr 07, 2017 at 01:29:27PM +0200, Juergen Gross wrote:
>> On 07/04/17 10:30, Oleksandr Andrushchenko wrote:
>>> From: Oleksandr Andrushchenko <oleksandr_andrushchenko@epam.com>
>>>
>>> The patch clarifies the protocol that is used by the PV keyboard
>>> drivers.
>>>
>>> Reviewed-by: Konrad Rzeszutek Wilk <konrad.wilk@oracle.com>
>>> Reviewed-by: Stefano Stabellini <sstabellini@kernel.org>
>>> Signed-off-by: Oleksandr Andrushchenko <oleksandr_andrushchenko@epam.com>
>> Acked-by: Juergen Gross <jgross@suse.com>
>>
>> I don't think its appropriate to keep the R-b:'s from the Xen
>> commits for the Linux kernel.
> <nods>
>
>> I can remove them when committing in case no V2 will be necessary.
Please, do me a favor
> Thank you!
>
> Oleksandr, thank you for doing this!
My pleasure
>> Added Boris.
>>
>>
>> Juergen
>>
>> _______________________________________________
>> Xen-devel mailing list
>> Xen-devel@lists.xen.org
>> https://lists.xen.org/xen-devel
Thank you,
Oleksandr
_______________________________________________
Xen-devel mailing list
Xen-devel@lists.xen.org
https://lists.xen.org/xen-devel
^ permalink raw reply [flat|nested] 41+ messages in thread
* Re: [For Linux 4/4] xen/displif: add ABI for para-virtual display
2017-04-07 8:30 ` [For Linux 4/4] xen/displif: add ABI for para-virtual display Oleksandr Andrushchenko
2017-04-07 11:31 ` [Xen-devel] " Juergen Gross
2017-04-07 11:31 ` Juergen Gross
@ 2017-04-07 13:50 ` Julien Grall
2017-04-07 14:02 ` Oleksandr Andrushchenko
2017-04-07 14:02 ` Oleksandr Andrushchenko
2017-04-07 13:50 ` Julien Grall
3 siblings, 2 replies; 41+ messages in thread
From: Julien Grall @ 2017-04-07 13:50 UTC (permalink / raw)
To: Oleksandr Andrushchenko, xen-devel
Cc: joculator, al1img, vlad.babchuk, andrii.anisov, olekstysh,
konrad.wilk, lars.kurth, sstabellini, linux-kernel,
Oleksandr Andrushchenko, Oleksandr Grytsov
Hi Oleksandr,
On 07/04/17 09:30, Oleksandr Andrushchenko wrote:
> +/*
> + ******************************************************************************
> + * Back to front events delivery
> + ******************************************************************************
> + * In order to deliver asynchronous events from back to front a shared page is
> + * allocated by front and its granted reference propagated to back via
> + * XenStore entries (evt-ring-ref/evt-event-channel).
> + * This page has a common header used by both front and back to synchronize
> + * access and control event's ring buffer, while back being a producer of the
> + * events and front being a consumer. The rest of the page after the header
> + * is used for event packets.
> + *
> + * Upon reception of an event(s) front may confirm its reception
> + * for either each event, group of events or none.
> + */
> +
> +struct xendispl_event_page {
> + uint32_t in_cons;
> + uint32_t in_prod;
> + uint8_t reserved[56];
> +};
> +
> +#define XENDISPL_EVENT_PAGE_SIZE 4096
This will be always the size of a grant (e.g xen page size), right? If
so, I would prefer if you use XEN_PAGE_SIZE to so we can easily update
Linux in the case Xen is using a different page size.
Note that the Front-end, Backend and Xen may use different page size, so
your drivers would need to cope with that. The current approach is to
always use the page granularity of Xen.
> +#define XENDISPL_IN_RING_OFFS (sizeof(struct xendispl_event_page))
> +#define XENDISPL_IN_RING_SIZE (XENDISPL_EVENT_PAGE_SIZE - XENDISPL_IN_RING_OFFS)
> +#define XENDISPL_IN_RING_LEN (XENDISPL_IN_RING_SIZE / sizeof(struct xendispl_evt))
> +#define XENDISPL_IN_RING(page) \
> + ((struct xendispl_evt *)((char *)(page) + XENDISPL_IN_RING_OFFS))
> +#define XENDISPL_IN_RING_REF(page, idx) \
> + (XENDISPL_IN_RING((page))[(idx) % XENDISPL_IN_RING_LEN])
> +
> +#endif /* __XEN_PUBLIC_IO_DISPLIF_H__ */
> +
> +/*
> + * Local variables:
> + * mode: C
> + * c-file-style: "BSD"
> + * c-basic-offset: 4
> + * tab-width: 4
> + * indent-tabs-mode: nil
> + * End:
> + */
>
Cheers,
--
Julien Grall
^ permalink raw reply [flat|nested] 41+ messages in thread
* Re: [For Linux 4/4] xen/displif: add ABI for para-virtual display
2017-04-07 8:30 ` [For Linux 4/4] xen/displif: add ABI for para-virtual display Oleksandr Andrushchenko
` (2 preceding siblings ...)
2017-04-07 13:50 ` Julien Grall
@ 2017-04-07 13:50 ` Julien Grall
3 siblings, 0 replies; 41+ messages in thread
From: Julien Grall @ 2017-04-07 13:50 UTC (permalink / raw)
To: Oleksandr Andrushchenko, xen-devel
Cc: lars.kurth, sstabellini, vlad.babchuk, Oleksandr Andrushchenko,
linux-kernel, andrii.anisov, olekstysh, al1img,
Oleksandr Grytsov, joculator
Hi Oleksandr,
On 07/04/17 09:30, Oleksandr Andrushchenko wrote:
> +/*
> + ******************************************************************************
> + * Back to front events delivery
> + ******************************************************************************
> + * In order to deliver asynchronous events from back to front a shared page is
> + * allocated by front and its granted reference propagated to back via
> + * XenStore entries (evt-ring-ref/evt-event-channel).
> + * This page has a common header used by both front and back to synchronize
> + * access and control event's ring buffer, while back being a producer of the
> + * events and front being a consumer. The rest of the page after the header
> + * is used for event packets.
> + *
> + * Upon reception of an event(s) front may confirm its reception
> + * for either each event, group of events or none.
> + */
> +
> +struct xendispl_event_page {
> + uint32_t in_cons;
> + uint32_t in_prod;
> + uint8_t reserved[56];
> +};
> +
> +#define XENDISPL_EVENT_PAGE_SIZE 4096
This will be always the size of a grant (e.g xen page size), right? If
so, I would prefer if you use XEN_PAGE_SIZE to so we can easily update
Linux in the case Xen is using a different page size.
Note that the Front-end, Backend and Xen may use different page size, so
your drivers would need to cope with that. The current approach is to
always use the page granularity of Xen.
> +#define XENDISPL_IN_RING_OFFS (sizeof(struct xendispl_event_page))
> +#define XENDISPL_IN_RING_SIZE (XENDISPL_EVENT_PAGE_SIZE - XENDISPL_IN_RING_OFFS)
> +#define XENDISPL_IN_RING_LEN (XENDISPL_IN_RING_SIZE / sizeof(struct xendispl_evt))
> +#define XENDISPL_IN_RING(page) \
> + ((struct xendispl_evt *)((char *)(page) + XENDISPL_IN_RING_OFFS))
> +#define XENDISPL_IN_RING_REF(page, idx) \
> + (XENDISPL_IN_RING((page))[(idx) % XENDISPL_IN_RING_LEN])
> +
> +#endif /* __XEN_PUBLIC_IO_DISPLIF_H__ */
> +
> +/*
> + * Local variables:
> + * mode: C
> + * c-file-style: "BSD"
> + * c-basic-offset: 4
> + * tab-width: 4
> + * indent-tabs-mode: nil
> + * End:
> + */
>
Cheers,
--
Julien Grall
_______________________________________________
Xen-devel mailing list
Xen-devel@lists.xen.org
https://lists.xen.org/xen-devel
^ permalink raw reply [flat|nested] 41+ messages in thread
* Re: [For Linux 4/4] xen/displif: add ABI for para-virtual display
2017-04-07 13:50 ` Julien Grall
2017-04-07 14:02 ` Oleksandr Andrushchenko
@ 2017-04-07 14:02 ` Oleksandr Andrushchenko
2017-04-07 16:36 ` Stefano Stabellini
` (3 more replies)
1 sibling, 4 replies; 41+ messages in thread
From: Oleksandr Andrushchenko @ 2017-04-07 14:02 UTC (permalink / raw)
To: Julien Grall, xen-devel
Cc: joculator, al1img, vlad.babchuk, andrii.anisov, olekstysh,
konrad.wilk, lars.kurth, sstabellini, linux-kernel,
Oleksandr Andrushchenko, Oleksandr Grytsov
Hi, Julien!
On 04/07/2017 04:50 PM, Julien Grall wrote:
> Hi Oleksandr,
>
> On 07/04/17 09:30, Oleksandr Andrushchenko wrote:
>> +/*
>> +
>> ******************************************************************************
>> + * Back to front events delivery
>> +
>> ******************************************************************************
>> + * In order to deliver asynchronous events from back to front a
>> shared page is
>> + * allocated by front and its granted reference propagated to back via
>> + * XenStore entries (evt-ring-ref/evt-event-channel).
>> + * This page has a common header used by both front and back to
>> synchronize
>> + * access and control event's ring buffer, while back being a
>> producer of the
>> + * events and front being a consumer. The rest of the page after the
>> header
>> + * is used for event packets.
>> + *
>> + * Upon reception of an event(s) front may confirm its reception
>> + * for either each event, group of events or none.
>> + */
>> +
>> +struct xendispl_event_page {
>> + uint32_t in_cons;
>> + uint32_t in_prod;
>> + uint8_t reserved[56];
>> +};
>> +
>> +#define XENDISPL_EVENT_PAGE_SIZE 4096
>
> This will be always the size of a grant (e.g xen page size), right? If
> so, I would prefer if you use XEN_PAGE_SIZE to so we can easily update
> Linux in the case Xen is using a different page size.
>
> Note that the Front-end, Backend and Xen may use different page size,
> so your drivers would need to cope with that. The current approach is
> to always use the page granularity of Xen.
>
Will the following satisfy the requirement?
#define XENDISPL_EVENT_PAGE_SIZE XEN_PAGE_SIZE
>> +#define XENDISPL_IN_RING_OFFS (sizeof(struct xendispl_event_page))
>> +#define XENDISPL_IN_RING_SIZE (XENDISPL_EVENT_PAGE_SIZE -
>> XENDISPL_IN_RING_OFFS)
>> +#define XENDISPL_IN_RING_LEN (XENDISPL_IN_RING_SIZE / sizeof(struct
>> xendispl_evt))
>> +#define XENDISPL_IN_RING(page) \
>> + ((struct xendispl_evt *)((char *)(page) + XENDISPL_IN_RING_OFFS))
>> +#define XENDISPL_IN_RING_REF(page, idx) \
>> + (XENDISPL_IN_RING((page))[(idx) % XENDISPL_IN_RING_LEN])
>> +
>> +#endif /* __XEN_PUBLIC_IO_DISPLIF_H__ */
>> +
>> +/*
>> + * Local variables:
>> + * mode: C
>> + * c-file-style: "BSD"
>> + * c-basic-offset: 4
>> + * tab-width: 4
>> + * indent-tabs-mode: nil
>> + * End:
>> + */
>>
>
> Cheers,
>
Thank you,
Oleksandr
^ permalink raw reply [flat|nested] 41+ messages in thread
* Re: [For Linux 4/4] xen/displif: add ABI for para-virtual display
2017-04-07 13:50 ` Julien Grall
@ 2017-04-07 14:02 ` Oleksandr Andrushchenko
2017-04-07 14:02 ` Oleksandr Andrushchenko
1 sibling, 0 replies; 41+ messages in thread
From: Oleksandr Andrushchenko @ 2017-04-07 14:02 UTC (permalink / raw)
To: Julien Grall, xen-devel
Cc: lars.kurth, sstabellini, vlad.babchuk, Oleksandr Andrushchenko,
linux-kernel, andrii.anisov, olekstysh, al1img,
Oleksandr Grytsov, joculator
Hi, Julien!
On 04/07/2017 04:50 PM, Julien Grall wrote:
> Hi Oleksandr,
>
> On 07/04/17 09:30, Oleksandr Andrushchenko wrote:
>> +/*
>> +
>> ******************************************************************************
>> + * Back to front events delivery
>> +
>> ******************************************************************************
>> + * In order to deliver asynchronous events from back to front a
>> shared page is
>> + * allocated by front and its granted reference propagated to back via
>> + * XenStore entries (evt-ring-ref/evt-event-channel).
>> + * This page has a common header used by both front and back to
>> synchronize
>> + * access and control event's ring buffer, while back being a
>> producer of the
>> + * events and front being a consumer. The rest of the page after the
>> header
>> + * is used for event packets.
>> + *
>> + * Upon reception of an event(s) front may confirm its reception
>> + * for either each event, group of events or none.
>> + */
>> +
>> +struct xendispl_event_page {
>> + uint32_t in_cons;
>> + uint32_t in_prod;
>> + uint8_t reserved[56];
>> +};
>> +
>> +#define XENDISPL_EVENT_PAGE_SIZE 4096
>
> This will be always the size of a grant (e.g xen page size), right? If
> so, I would prefer if you use XEN_PAGE_SIZE to so we can easily update
> Linux in the case Xen is using a different page size.
>
> Note that the Front-end, Backend and Xen may use different page size,
> so your drivers would need to cope with that. The current approach is
> to always use the page granularity of Xen.
>
Will the following satisfy the requirement?
#define XENDISPL_EVENT_PAGE_SIZE XEN_PAGE_SIZE
>> +#define XENDISPL_IN_RING_OFFS (sizeof(struct xendispl_event_page))
>> +#define XENDISPL_IN_RING_SIZE (XENDISPL_EVENT_PAGE_SIZE -
>> XENDISPL_IN_RING_OFFS)
>> +#define XENDISPL_IN_RING_LEN (XENDISPL_IN_RING_SIZE / sizeof(struct
>> xendispl_evt))
>> +#define XENDISPL_IN_RING(page) \
>> + ((struct xendispl_evt *)((char *)(page) + XENDISPL_IN_RING_OFFS))
>> +#define XENDISPL_IN_RING_REF(page, idx) \
>> + (XENDISPL_IN_RING((page))[(idx) % XENDISPL_IN_RING_LEN])
>> +
>> +#endif /* __XEN_PUBLIC_IO_DISPLIF_H__ */
>> +
>> +/*
>> + * Local variables:
>> + * mode: C
>> + * c-file-style: "BSD"
>> + * c-basic-offset: 4
>> + * tab-width: 4
>> + * indent-tabs-mode: nil
>> + * End:
>> + */
>>
>
> Cheers,
>
Thank you,
Oleksandr
_______________________________________________
Xen-devel mailing list
Xen-devel@lists.xen.org
https://lists.xen.org/xen-devel
^ permalink raw reply [flat|nested] 41+ messages in thread
* Re: [For Linux 4/4] xen/displif: add ABI for para-virtual display
2017-04-07 14:02 ` Oleksandr Andrushchenko
@ 2017-04-07 16:36 ` Stefano Stabellini
2017-04-07 17:43 ` Oleksandr Andrushchenko
2017-04-07 17:43 ` Oleksandr Andrushchenko
2017-04-07 16:36 ` Stefano Stabellini
` (2 subsequent siblings)
3 siblings, 2 replies; 41+ messages in thread
From: Stefano Stabellini @ 2017-04-07 16:36 UTC (permalink / raw)
To: Oleksandr Andrushchenko
Cc: Julien Grall, xen-devel, joculator, al1img, vlad.babchuk,
andrii.anisov, olekstysh, konrad.wilk, lars.kurth, sstabellini,
linux-kernel, Oleksandr Andrushchenko, Oleksandr Grytsov
On Fri, 7 Apr 2017, Oleksandr Andrushchenko wrote:
> Hi, Julien!
>
> On 04/07/2017 04:50 PM, Julien Grall wrote:
> > Hi Oleksandr,
> >
> > On 07/04/17 09:30, Oleksandr Andrushchenko wrote:
> > > +/*
> > > +
> > > ******************************************************************************
> > > + * Back to front events delivery
> > > +
> > > ******************************************************************************
> > > + * In order to deliver asynchronous events from back to front a shared
> > > page is
> > > + * allocated by front and its granted reference propagated to back via
> > > + * XenStore entries (evt-ring-ref/evt-event-channel).
> > > + * This page has a common header used by both front and back to
> > > synchronize
> > > + * access and control event's ring buffer, while back being a producer of
> > > the
> > > + * events and front being a consumer. The rest of the page after the
> > > header
> > > + * is used for event packets.
> > > + *
> > > + * Upon reception of an event(s) front may confirm its reception
> > > + * for either each event, group of events or none.
> > > + */
> > > +
> > > +struct xendispl_event_page {
> > > + uint32_t in_cons;
> > > + uint32_t in_prod;
> > > + uint8_t reserved[56];
> > > +};
> > > +
> > > +#define XENDISPL_EVENT_PAGE_SIZE 4096
> >
> > This will be always the size of a grant (e.g xen page size), right? If so, I
> > would prefer if you use XEN_PAGE_SIZE to so we can easily update Linux in
> > the case Xen is using a different page size.
> >
> > Note that the Front-end, Backend and Xen may use different page size, so
> > your drivers would need to cope with that. The current approach is to always
> > use the page granularity of Xen.
> >
> Will the following satisfy the requirement?
> #define XENDISPL_EVENT_PAGE_SIZE XEN_PAGE_SIZE
It is already defined in include/xen/page.h
> > > +#define XENDISPL_IN_RING_OFFS (sizeof(struct xendispl_event_page))
> > > +#define XENDISPL_IN_RING_SIZE (XENDISPL_EVENT_PAGE_SIZE -
> > > XENDISPL_IN_RING_OFFS)
> > > +#define XENDISPL_IN_RING_LEN (XENDISPL_IN_RING_SIZE / sizeof(struct
> > > xendispl_evt))
> > > +#define XENDISPL_IN_RING(page) \
> > > + ((struct xendispl_evt *)((char *)(page) + XENDISPL_IN_RING_OFFS))
> > > +#define XENDISPL_IN_RING_REF(page, idx) \
> > > + (XENDISPL_IN_RING((page))[(idx) % XENDISPL_IN_RING_LEN])
> > > +
> > > +#endif /* __XEN_PUBLIC_IO_DISPLIF_H__ */
> > > +
> > > +/*
> > > + * Local variables:
> > > + * mode: C
> > > + * c-file-style: "BSD"
> > > + * c-basic-offset: 4
> > > + * tab-width: 4
> > > + * indent-tabs-mode: nil
> > > + * End:
> > > + */
> > >
> >
> > Cheers,
> >
> Thank you,
> Oleksandr
>
^ permalink raw reply [flat|nested] 41+ messages in thread
* Re: [For Linux 4/4] xen/displif: add ABI for para-virtual display
2017-04-07 14:02 ` Oleksandr Andrushchenko
2017-04-07 16:36 ` Stefano Stabellini
@ 2017-04-07 16:36 ` Stefano Stabellini
2017-04-10 6:03 ` Juergen Gross
2017-04-10 6:03 ` Juergen Gross
3 siblings, 0 replies; 41+ messages in thread
From: Stefano Stabellini @ 2017-04-07 16:36 UTC (permalink / raw)
To: Oleksandr Andrushchenko
Cc: lars.kurth, sstabellini, vlad.babchuk, linux-kernel,
Oleksandr Andrushchenko, al1img, andrii.anisov, olekstysh,
Julien Grall, Oleksandr Grytsov, xen-devel, joculator
On Fri, 7 Apr 2017, Oleksandr Andrushchenko wrote:
> Hi, Julien!
>
> On 04/07/2017 04:50 PM, Julien Grall wrote:
> > Hi Oleksandr,
> >
> > On 07/04/17 09:30, Oleksandr Andrushchenko wrote:
> > > +/*
> > > +
> > > ******************************************************************************
> > > + * Back to front events delivery
> > > +
> > > ******************************************************************************
> > > + * In order to deliver asynchronous events from back to front a shared
> > > page is
> > > + * allocated by front and its granted reference propagated to back via
> > > + * XenStore entries (evt-ring-ref/evt-event-channel).
> > > + * This page has a common header used by both front and back to
> > > synchronize
> > > + * access and control event's ring buffer, while back being a producer of
> > > the
> > > + * events and front being a consumer. The rest of the page after the
> > > header
> > > + * is used for event packets.
> > > + *
> > > + * Upon reception of an event(s) front may confirm its reception
> > > + * for either each event, group of events or none.
> > > + */
> > > +
> > > +struct xendispl_event_page {
> > > + uint32_t in_cons;
> > > + uint32_t in_prod;
> > > + uint8_t reserved[56];
> > > +};
> > > +
> > > +#define XENDISPL_EVENT_PAGE_SIZE 4096
> >
> > This will be always the size of a grant (e.g xen page size), right? If so, I
> > would prefer if you use XEN_PAGE_SIZE to so we can easily update Linux in
> > the case Xen is using a different page size.
> >
> > Note that the Front-end, Backend and Xen may use different page size, so
> > your drivers would need to cope with that. The current approach is to always
> > use the page granularity of Xen.
> >
> Will the following satisfy the requirement?
> #define XENDISPL_EVENT_PAGE_SIZE XEN_PAGE_SIZE
It is already defined in include/xen/page.h
> > > +#define XENDISPL_IN_RING_OFFS (sizeof(struct xendispl_event_page))
> > > +#define XENDISPL_IN_RING_SIZE (XENDISPL_EVENT_PAGE_SIZE -
> > > XENDISPL_IN_RING_OFFS)
> > > +#define XENDISPL_IN_RING_LEN (XENDISPL_IN_RING_SIZE / sizeof(struct
> > > xendispl_evt))
> > > +#define XENDISPL_IN_RING(page) \
> > > + ((struct xendispl_evt *)((char *)(page) + XENDISPL_IN_RING_OFFS))
> > > +#define XENDISPL_IN_RING_REF(page, idx) \
> > > + (XENDISPL_IN_RING((page))[(idx) % XENDISPL_IN_RING_LEN])
> > > +
> > > +#endif /* __XEN_PUBLIC_IO_DISPLIF_H__ */
> > > +
> > > +/*
> > > + * Local variables:
> > > + * mode: C
> > > + * c-file-style: "BSD"
> > > + * c-basic-offset: 4
> > > + * tab-width: 4
> > > + * indent-tabs-mode: nil
> > > + * End:
> > > + */
> > >
> >
> > Cheers,
> >
> Thank you,
> Oleksandr
>
_______________________________________________
Xen-devel mailing list
Xen-devel@lists.xen.org
https://lists.xen.org/xen-devel
^ permalink raw reply [flat|nested] 41+ messages in thread
* Re: [For Linux 4/4] xen/displif: add ABI for para-virtual display
2017-04-07 16:36 ` Stefano Stabellini
2017-04-07 17:43 ` Oleksandr Andrushchenko
@ 2017-04-07 17:43 ` Oleksandr Andrushchenko
2017-04-07 17:50 ` Stefano Stabellini
2017-04-07 17:50 ` Stefano Stabellini
1 sibling, 2 replies; 41+ messages in thread
From: Oleksandr Andrushchenko @ 2017-04-07 17:43 UTC (permalink / raw)
To: Stefano Stabellini
Cc: Julien Grall, xen-devel, joculator, al1img, vlad.babchuk,
andrii.anisov, olekstysh, konrad.wilk, lars.kurth, linux-kernel,
Oleksandr Andrushchenko, Oleksandr Grytsov
On 04/07/2017 07:36 PM, Stefano Stabellini wrote:
> On Fri, 7 Apr 2017, Oleksandr Andrushchenko wrote:
>> Hi, Julien!
>>
>> On 04/07/2017 04:50 PM, Julien Grall wrote:
>>> Hi Oleksandr,
>>>
>>> On 07/04/17 09:30, Oleksandr Andrushchenko wrote:
>>>> +/*
>>>> +
>>>> ******************************************************************************
>>>> + * Back to front events delivery
>>>> +
>>>> ******************************************************************************
>>>> + * In order to deliver asynchronous events from back to front a shared
>>>> page is
>>>> + * allocated by front and its granted reference propagated to back via
>>>> + * XenStore entries (evt-ring-ref/evt-event-channel).
>>>> + * This page has a common header used by both front and back to
>>>> synchronize
>>>> + * access and control event's ring buffer, while back being a producer of
>>>> the
>>>> + * events and front being a consumer. The rest of the page after the
>>>> header
>>>> + * is used for event packets.
>>>> + *
>>>> + * Upon reception of an event(s) front may confirm its reception
>>>> + * for either each event, group of events or none.
>>>> + */
>>>> +
>>>> +struct xendispl_event_page {
>>>> + uint32_t in_cons;
>>>> + uint32_t in_prod;
>>>> + uint8_t reserved[56];
>>>> +};
>>>> +
>>>> +#define XENDISPL_EVENT_PAGE_SIZE 4096
>>> This will be always the size of a grant (e.g xen page size), right? If so, I
>>> would prefer if you use XEN_PAGE_SIZE to so we can easily update Linux in
>>> the case Xen is using a different page size.
>>>
>>> Note that the Front-end, Backend and Xen may use different page size, so
>>> your drivers would need to cope with that. The current approach is to always
>>> use the page granularity of Xen.
>>>
>> Will the following satisfy the requirement?
>> #define XENDISPL_EVENT_PAGE_SIZE XEN_PAGE_SIZE
> It is already defined in include/xen/page.h
Thank you, Stefano, but I meant if the change below
will fit the requirement, e.g. what Julien means is:
-#define XENDISPL_EVENT_PAGE_SIZE 4096
+#define XENDISPL_EVENT_PAGE_SIZE XEN_PAGE_SIZE
Julien, does this change work for you?
>
>>>> +#define XENDISPL_IN_RING_OFFS (sizeof(struct xendispl_event_page))
>>>> +#define XENDISPL_IN_RING_SIZE (XENDISPL_EVENT_PAGE_SIZE -
>>>> XENDISPL_IN_RING_OFFS)
>>>> +#define XENDISPL_IN_RING_LEN (XENDISPL_IN_RING_SIZE / sizeof(struct
>>>> xendispl_evt))
>>>> +#define XENDISPL_IN_RING(page) \
>>>> + ((struct xendispl_evt *)((char *)(page) + XENDISPL_IN_RING_OFFS))
>>>> +#define XENDISPL_IN_RING_REF(page, idx) \
>>>> + (XENDISPL_IN_RING((page))[(idx) % XENDISPL_IN_RING_LEN])
>>>> +
>>>> +#endif /* __XEN_PUBLIC_IO_DISPLIF_H__ */
>>>> +
>>>> +/*
>>>> + * Local variables:
>>>> + * mode: C
>>>> + * c-file-style: "BSD"
>>>> + * c-basic-offset: 4
>>>> + * tab-width: 4
>>>> + * indent-tabs-mode: nil
>>>> + * End:
>>>> + */
>>>>
>>> Cheers,
>>>
>> Thank you,
>> Oleksandr
>>
Thank you,
Oleksandr
^ permalink raw reply [flat|nested] 41+ messages in thread
* Re: [For Linux 4/4] xen/displif: add ABI for para-virtual display
2017-04-07 16:36 ` Stefano Stabellini
@ 2017-04-07 17:43 ` Oleksandr Andrushchenko
2017-04-07 17:43 ` Oleksandr Andrushchenko
1 sibling, 0 replies; 41+ messages in thread
From: Oleksandr Andrushchenko @ 2017-04-07 17:43 UTC (permalink / raw)
To: Stefano Stabellini
Cc: lars.kurth, vlad.babchuk, linux-kernel, Oleksandr Andrushchenko,
al1img, andrii.anisov, olekstysh, Julien Grall,
Oleksandr Grytsov, xen-devel, joculator
On 04/07/2017 07:36 PM, Stefano Stabellini wrote:
> On Fri, 7 Apr 2017, Oleksandr Andrushchenko wrote:
>> Hi, Julien!
>>
>> On 04/07/2017 04:50 PM, Julien Grall wrote:
>>> Hi Oleksandr,
>>>
>>> On 07/04/17 09:30, Oleksandr Andrushchenko wrote:
>>>> +/*
>>>> +
>>>> ******************************************************************************
>>>> + * Back to front events delivery
>>>> +
>>>> ******************************************************************************
>>>> + * In order to deliver asynchronous events from back to front a shared
>>>> page is
>>>> + * allocated by front and its granted reference propagated to back via
>>>> + * XenStore entries (evt-ring-ref/evt-event-channel).
>>>> + * This page has a common header used by both front and back to
>>>> synchronize
>>>> + * access and control event's ring buffer, while back being a producer of
>>>> the
>>>> + * events and front being a consumer. The rest of the page after the
>>>> header
>>>> + * is used for event packets.
>>>> + *
>>>> + * Upon reception of an event(s) front may confirm its reception
>>>> + * for either each event, group of events or none.
>>>> + */
>>>> +
>>>> +struct xendispl_event_page {
>>>> + uint32_t in_cons;
>>>> + uint32_t in_prod;
>>>> + uint8_t reserved[56];
>>>> +};
>>>> +
>>>> +#define XENDISPL_EVENT_PAGE_SIZE 4096
>>> This will be always the size of a grant (e.g xen page size), right? If so, I
>>> would prefer if you use XEN_PAGE_SIZE to so we can easily update Linux in
>>> the case Xen is using a different page size.
>>>
>>> Note that the Front-end, Backend and Xen may use different page size, so
>>> your drivers would need to cope with that. The current approach is to always
>>> use the page granularity of Xen.
>>>
>> Will the following satisfy the requirement?
>> #define XENDISPL_EVENT_PAGE_SIZE XEN_PAGE_SIZE
> It is already defined in include/xen/page.h
Thank you, Stefano, but I meant if the change below
will fit the requirement, e.g. what Julien means is:
-#define XENDISPL_EVENT_PAGE_SIZE 4096
+#define XENDISPL_EVENT_PAGE_SIZE XEN_PAGE_SIZE
Julien, does this change work for you?
>
>>>> +#define XENDISPL_IN_RING_OFFS (sizeof(struct xendispl_event_page))
>>>> +#define XENDISPL_IN_RING_SIZE (XENDISPL_EVENT_PAGE_SIZE -
>>>> XENDISPL_IN_RING_OFFS)
>>>> +#define XENDISPL_IN_RING_LEN (XENDISPL_IN_RING_SIZE / sizeof(struct
>>>> xendispl_evt))
>>>> +#define XENDISPL_IN_RING(page) \
>>>> + ((struct xendispl_evt *)((char *)(page) + XENDISPL_IN_RING_OFFS))
>>>> +#define XENDISPL_IN_RING_REF(page, idx) \
>>>> + (XENDISPL_IN_RING((page))[(idx) % XENDISPL_IN_RING_LEN])
>>>> +
>>>> +#endif /* __XEN_PUBLIC_IO_DISPLIF_H__ */
>>>> +
>>>> +/*
>>>> + * Local variables:
>>>> + * mode: C
>>>> + * c-file-style: "BSD"
>>>> + * c-basic-offset: 4
>>>> + * tab-width: 4
>>>> + * indent-tabs-mode: nil
>>>> + * End:
>>>> + */
>>>>
>>> Cheers,
>>>
>> Thank you,
>> Oleksandr
>>
Thank you,
Oleksandr
_______________________________________________
Xen-devel mailing list
Xen-devel@lists.xen.org
https://lists.xen.org/xen-devel
^ permalink raw reply [flat|nested] 41+ messages in thread
* Re: [For Linux 4/4] xen/displif: add ABI for para-virtual display
2017-04-07 17:43 ` Oleksandr Andrushchenko
2017-04-07 17:50 ` Stefano Stabellini
@ 2017-04-07 17:50 ` Stefano Stabellini
2017-04-07 17:51 ` Oleksandr Andrushchenko
2017-04-07 17:51 ` Oleksandr Andrushchenko
1 sibling, 2 replies; 41+ messages in thread
From: Stefano Stabellini @ 2017-04-07 17:50 UTC (permalink / raw)
To: Oleksandr Andrushchenko
Cc: Stefano Stabellini, Julien Grall, xen-devel, joculator, al1img,
vlad.babchuk, andrii.anisov, olekstysh, konrad.wilk, lars.kurth,
linux-kernel, Oleksandr Andrushchenko, Oleksandr Grytsov
On Fri, 7 Apr 2017, Oleksandr Andrushchenko wrote:
> On 04/07/2017 07:36 PM, Stefano Stabellini wrote:
> > On Fri, 7 Apr 2017, Oleksandr Andrushchenko wrote:
> > > Hi, Julien!
> > >
> > > On 04/07/2017 04:50 PM, Julien Grall wrote:
> > > > Hi Oleksandr,
> > > >
> > > > On 07/04/17 09:30, Oleksandr Andrushchenko wrote:
> > > > > +/*
> > > > > +
> > > > > ******************************************************************************
> > > > > + * Back to front events delivery
> > > > > +
> > > > > ******************************************************************************
> > > > > + * In order to deliver asynchronous events from back to front a
> > > > > shared
> > > > > page is
> > > > > + * allocated by front and its granted reference propagated to back
> > > > > via
> > > > > + * XenStore entries (evt-ring-ref/evt-event-channel).
> > > > > + * This page has a common header used by both front and back to
> > > > > synchronize
> > > > > + * access and control event's ring buffer, while back being a
> > > > > producer of
> > > > > the
> > > > > + * events and front being a consumer. The rest of the page after the
> > > > > header
> > > > > + * is used for event packets.
> > > > > + *
> > > > > + * Upon reception of an event(s) front may confirm its reception
> > > > > + * for either each event, group of events or none.
> > > > > + */
> > > > > +
> > > > > +struct xendispl_event_page {
> > > > > + uint32_t in_cons;
> > > > > + uint32_t in_prod;
> > > > > + uint8_t reserved[56];
> > > > > +};
> > > > > +
> > > > > +#define XENDISPL_EVENT_PAGE_SIZE 4096
> > > > This will be always the size of a grant (e.g xen page size), right? If
> > > > so, I
> > > > would prefer if you use XEN_PAGE_SIZE to so we can easily update Linux
> > > > in
> > > > the case Xen is using a different page size.
> > > >
> > > > Note that the Front-end, Backend and Xen may use different page size, so
> > > > your drivers would need to cope with that. The current approach is to
> > > > always
> > > > use the page granularity of Xen.
> > > >
> > > Will the following satisfy the requirement?
> > > #define XENDISPL_EVENT_PAGE_SIZE XEN_PAGE_SIZE
> > It is already defined in include/xen/page.h
> Thank you, Stefano, but I meant if the change below
> will fit the requirement, e.g. what Julien means is:
> -#define XENDISPL_EVENT_PAGE_SIZE 4096
> +#define XENDISPL_EVENT_PAGE_SIZE XEN_PAGE_SIZE
> Julien, does this change work for you?
Sure, I think that would be OK. Also using XEN_PAGE_SIZE directly would
be OK.
> > > > > +#define XENDISPL_IN_RING_OFFS (sizeof(struct xendispl_event_page))
> > > > > +#define XENDISPL_IN_RING_SIZE (XENDISPL_EVENT_PAGE_SIZE -
> > > > > XENDISPL_IN_RING_OFFS)
> > > > > +#define XENDISPL_IN_RING_LEN (XENDISPL_IN_RING_SIZE / sizeof(struct
> > > > > xendispl_evt))
> > > > > +#define XENDISPL_IN_RING(page) \
> > > > > + ((struct xendispl_evt *)((char *)(page) + XENDISPL_IN_RING_OFFS))
> > > > > +#define XENDISPL_IN_RING_REF(page, idx) \
> > > > > + (XENDISPL_IN_RING((page))[(idx) % XENDISPL_IN_RING_LEN])
> > > > > +
> > > > > +#endif /* __XEN_PUBLIC_IO_DISPLIF_H__ */
> > > > > +
> > > > > +/*
> > > > > + * Local variables:
> > > > > + * mode: C
> > > > > + * c-file-style: "BSD"
> > > > > + * c-basic-offset: 4
> > > > > + * tab-width: 4
> > > > > + * indent-tabs-mode: nil
> > > > > + * End:
> > > > > + */
> > > > >
> > > > Cheers,
> > > >
> > > Thank you,
> > > Oleksandr
> > >
> Thank you,
> Oleksandr
>
^ permalink raw reply [flat|nested] 41+ messages in thread
* Re: [For Linux 4/4] xen/displif: add ABI for para-virtual display
2017-04-07 17:43 ` Oleksandr Andrushchenko
@ 2017-04-07 17:50 ` Stefano Stabellini
2017-04-07 17:50 ` Stefano Stabellini
1 sibling, 0 replies; 41+ messages in thread
From: Stefano Stabellini @ 2017-04-07 17:50 UTC (permalink / raw)
To: Oleksandr Andrushchenko
Cc: lars.kurth, Stefano Stabellini, vlad.babchuk, linux-kernel,
Oleksandr Andrushchenko, al1img, andrii.anisov, olekstysh,
Julien Grall, Oleksandr Grytsov, xen-devel, joculator
On Fri, 7 Apr 2017, Oleksandr Andrushchenko wrote:
> On 04/07/2017 07:36 PM, Stefano Stabellini wrote:
> > On Fri, 7 Apr 2017, Oleksandr Andrushchenko wrote:
> > > Hi, Julien!
> > >
> > > On 04/07/2017 04:50 PM, Julien Grall wrote:
> > > > Hi Oleksandr,
> > > >
> > > > On 07/04/17 09:30, Oleksandr Andrushchenko wrote:
> > > > > +/*
> > > > > +
> > > > > ******************************************************************************
> > > > > + * Back to front events delivery
> > > > > +
> > > > > ******************************************************************************
> > > > > + * In order to deliver asynchronous events from back to front a
> > > > > shared
> > > > > page is
> > > > > + * allocated by front and its granted reference propagated to back
> > > > > via
> > > > > + * XenStore entries (evt-ring-ref/evt-event-channel).
> > > > > + * This page has a common header used by both front and back to
> > > > > synchronize
> > > > > + * access and control event's ring buffer, while back being a
> > > > > producer of
> > > > > the
> > > > > + * events and front being a consumer. The rest of the page after the
> > > > > header
> > > > > + * is used for event packets.
> > > > > + *
> > > > > + * Upon reception of an event(s) front may confirm its reception
> > > > > + * for either each event, group of events or none.
> > > > > + */
> > > > > +
> > > > > +struct xendispl_event_page {
> > > > > + uint32_t in_cons;
> > > > > + uint32_t in_prod;
> > > > > + uint8_t reserved[56];
> > > > > +};
> > > > > +
> > > > > +#define XENDISPL_EVENT_PAGE_SIZE 4096
> > > > This will be always the size of a grant (e.g xen page size), right? If
> > > > so, I
> > > > would prefer if you use XEN_PAGE_SIZE to so we can easily update Linux
> > > > in
> > > > the case Xen is using a different page size.
> > > >
> > > > Note that the Front-end, Backend and Xen may use different page size, so
> > > > your drivers would need to cope with that. The current approach is to
> > > > always
> > > > use the page granularity of Xen.
> > > >
> > > Will the following satisfy the requirement?
> > > #define XENDISPL_EVENT_PAGE_SIZE XEN_PAGE_SIZE
> > It is already defined in include/xen/page.h
> Thank you, Stefano, but I meant if the change below
> will fit the requirement, e.g. what Julien means is:
> -#define XENDISPL_EVENT_PAGE_SIZE 4096
> +#define XENDISPL_EVENT_PAGE_SIZE XEN_PAGE_SIZE
> Julien, does this change work for you?
Sure, I think that would be OK. Also using XEN_PAGE_SIZE directly would
be OK.
> > > > > +#define XENDISPL_IN_RING_OFFS (sizeof(struct xendispl_event_page))
> > > > > +#define XENDISPL_IN_RING_SIZE (XENDISPL_EVENT_PAGE_SIZE -
> > > > > XENDISPL_IN_RING_OFFS)
> > > > > +#define XENDISPL_IN_RING_LEN (XENDISPL_IN_RING_SIZE / sizeof(struct
> > > > > xendispl_evt))
> > > > > +#define XENDISPL_IN_RING(page) \
> > > > > + ((struct xendispl_evt *)((char *)(page) + XENDISPL_IN_RING_OFFS))
> > > > > +#define XENDISPL_IN_RING_REF(page, idx) \
> > > > > + (XENDISPL_IN_RING((page))[(idx) % XENDISPL_IN_RING_LEN])
> > > > > +
> > > > > +#endif /* __XEN_PUBLIC_IO_DISPLIF_H__ */
> > > > > +
> > > > > +/*
> > > > > + * Local variables:
> > > > > + * mode: C
> > > > > + * c-file-style: "BSD"
> > > > > + * c-basic-offset: 4
> > > > > + * tab-width: 4
> > > > > + * indent-tabs-mode: nil
> > > > > + * End:
> > > > > + */
> > > > >
> > > > Cheers,
> > > >
> > > Thank you,
> > > Oleksandr
> > >
> Thank you,
> Oleksandr
>
_______________________________________________
Xen-devel mailing list
Xen-devel@lists.xen.org
https://lists.xen.org/xen-devel
^ permalink raw reply [flat|nested] 41+ messages in thread
* Re: [For Linux 4/4] xen/displif: add ABI for para-virtual display
2017-04-07 17:50 ` Stefano Stabellini
@ 2017-04-07 17:51 ` Oleksandr Andrushchenko
2017-04-07 17:51 ` Oleksandr Andrushchenko
1 sibling, 0 replies; 41+ messages in thread
From: Oleksandr Andrushchenko @ 2017-04-07 17:51 UTC (permalink / raw)
To: Stefano Stabellini
Cc: Julien Grall, xen-devel, joculator, al1img, vlad.babchuk,
andrii.anisov, olekstysh, konrad.wilk, lars.kurth, linux-kernel,
Oleksandr Andrushchenko, Oleksandr Grytsov
On 04/07/2017 08:50 PM, Stefano Stabellini wrote:
> On Fri, 7 Apr 2017, Oleksandr Andrushchenko wrote:
>> On 04/07/2017 07:36 PM, Stefano Stabellini wrote:
>>> On Fri, 7 Apr 2017, Oleksandr Andrushchenko wrote:
>>>> Hi, Julien!
>>>>
>>>> On 04/07/2017 04:50 PM, Julien Grall wrote:
>>>>> Hi Oleksandr,
>>>>>
>>>>> On 07/04/17 09:30, Oleksandr Andrushchenko wrote:
>>>>>> +/*
>>>>>> +
>>>>>> ******************************************************************************
>>>>>> + * Back to front events delivery
>>>>>> +
>>>>>> ******************************************************************************
>>>>>> + * In order to deliver asynchronous events from back to front a
>>>>>> shared
>>>>>> page is
>>>>>> + * allocated by front and its granted reference propagated to back
>>>>>> via
>>>>>> + * XenStore entries (evt-ring-ref/evt-event-channel).
>>>>>> + * This page has a common header used by both front and back to
>>>>>> synchronize
>>>>>> + * access and control event's ring buffer, while back being a
>>>>>> producer of
>>>>>> the
>>>>>> + * events and front being a consumer. The rest of the page after the
>>>>>> header
>>>>>> + * is used for event packets.
>>>>>> + *
>>>>>> + * Upon reception of an event(s) front may confirm its reception
>>>>>> + * for either each event, group of events or none.
>>>>>> + */
>>>>>> +
>>>>>> +struct xendispl_event_page {
>>>>>> + uint32_t in_cons;
>>>>>> + uint32_t in_prod;
>>>>>> + uint8_t reserved[56];
>>>>>> +};
>>>>>> +
>>>>>> +#define XENDISPL_EVENT_PAGE_SIZE 4096
>>>>> This will be always the size of a grant (e.g xen page size), right? If
>>>>> so, I
>>>>> would prefer if you use XEN_PAGE_SIZE to so we can easily update Linux
>>>>> in
>>>>> the case Xen is using a different page size.
>>>>>
>>>>> Note that the Front-end, Backend and Xen may use different page size, so
>>>>> your drivers would need to cope with that. The current approach is to
>>>>> always
>>>>> use the page granularity of Xen.
>>>>>
>>>> Will the following satisfy the requirement?
>>>> #define XENDISPL_EVENT_PAGE_SIZE XEN_PAGE_SIZE
>>> It is already defined in include/xen/page.h
>> Thank you, Stefano, but I meant if the change below
>> will fit the requirement, e.g. what Julien means is:
>> -#define XENDISPL_EVENT_PAGE_SIZE 4096
>> +#define XENDISPL_EVENT_PAGE_SIZE XEN_PAGE_SIZE
>> Julien, does this change work for you?
> Sure, I think that would be OK. Also using XEN_PAGE_SIZE directly would
> be OK.
>
Great, thanks
>>>>>> +#define XENDISPL_IN_RING_OFFS (sizeof(struct xendispl_event_page))
>>>>>> +#define XENDISPL_IN_RING_SIZE (XENDISPL_EVENT_PAGE_SIZE -
>>>>>> XENDISPL_IN_RING_OFFS)
>>>>>> +#define XENDISPL_IN_RING_LEN (XENDISPL_IN_RING_SIZE / sizeof(struct
>>>>>> xendispl_evt))
>>>>>> +#define XENDISPL_IN_RING(page) \
>>>>>> + ((struct xendispl_evt *)((char *)(page) + XENDISPL_IN_RING_OFFS))
>>>>>> +#define XENDISPL_IN_RING_REF(page, idx) \
>>>>>> + (XENDISPL_IN_RING((page))[(idx) % XENDISPL_IN_RING_LEN])
>>>>>> +
>>>>>> +#endif /* __XEN_PUBLIC_IO_DISPLIF_H__ */
>>>>>> +
>>>>>> +/*
>>>>>> + * Local variables:
>>>>>> + * mode: C
>>>>>> + * c-file-style: "BSD"
>>>>>> + * c-basic-offset: 4
>>>>>> + * tab-width: 4
>>>>>> + * indent-tabs-mode: nil
>>>>>> + * End:
>>>>>> + */
>>>>>>
>>>>> Cheers,
>>>>>
>>>> Thank you,
>>>> Oleksandr
>>>>
>> Thank you,
>> Oleksandr
>>
^ permalink raw reply [flat|nested] 41+ messages in thread
* Re: [For Linux 4/4] xen/displif: add ABI for para-virtual display
2017-04-07 17:50 ` Stefano Stabellini
2017-04-07 17:51 ` Oleksandr Andrushchenko
@ 2017-04-07 17:51 ` Oleksandr Andrushchenko
1 sibling, 0 replies; 41+ messages in thread
From: Oleksandr Andrushchenko @ 2017-04-07 17:51 UTC (permalink / raw)
To: Stefano Stabellini
Cc: lars.kurth, vlad.babchuk, linux-kernel, Oleksandr Andrushchenko,
al1img, andrii.anisov, olekstysh, Julien Grall,
Oleksandr Grytsov, xen-devel, joculator
On 04/07/2017 08:50 PM, Stefano Stabellini wrote:
> On Fri, 7 Apr 2017, Oleksandr Andrushchenko wrote:
>> On 04/07/2017 07:36 PM, Stefano Stabellini wrote:
>>> On Fri, 7 Apr 2017, Oleksandr Andrushchenko wrote:
>>>> Hi, Julien!
>>>>
>>>> On 04/07/2017 04:50 PM, Julien Grall wrote:
>>>>> Hi Oleksandr,
>>>>>
>>>>> On 07/04/17 09:30, Oleksandr Andrushchenko wrote:
>>>>>> +/*
>>>>>> +
>>>>>> ******************************************************************************
>>>>>> + * Back to front events delivery
>>>>>> +
>>>>>> ******************************************************************************
>>>>>> + * In order to deliver asynchronous events from back to front a
>>>>>> shared
>>>>>> page is
>>>>>> + * allocated by front and its granted reference propagated to back
>>>>>> via
>>>>>> + * XenStore entries (evt-ring-ref/evt-event-channel).
>>>>>> + * This page has a common header used by both front and back to
>>>>>> synchronize
>>>>>> + * access and control event's ring buffer, while back being a
>>>>>> producer of
>>>>>> the
>>>>>> + * events and front being a consumer. The rest of the page after the
>>>>>> header
>>>>>> + * is used for event packets.
>>>>>> + *
>>>>>> + * Upon reception of an event(s) front may confirm its reception
>>>>>> + * for either each event, group of events or none.
>>>>>> + */
>>>>>> +
>>>>>> +struct xendispl_event_page {
>>>>>> + uint32_t in_cons;
>>>>>> + uint32_t in_prod;
>>>>>> + uint8_t reserved[56];
>>>>>> +};
>>>>>> +
>>>>>> +#define XENDISPL_EVENT_PAGE_SIZE 4096
>>>>> This will be always the size of a grant (e.g xen page size), right? If
>>>>> so, I
>>>>> would prefer if you use XEN_PAGE_SIZE to so we can easily update Linux
>>>>> in
>>>>> the case Xen is using a different page size.
>>>>>
>>>>> Note that the Front-end, Backend and Xen may use different page size, so
>>>>> your drivers would need to cope with that. The current approach is to
>>>>> always
>>>>> use the page granularity of Xen.
>>>>>
>>>> Will the following satisfy the requirement?
>>>> #define XENDISPL_EVENT_PAGE_SIZE XEN_PAGE_SIZE
>>> It is already defined in include/xen/page.h
>> Thank you, Stefano, but I meant if the change below
>> will fit the requirement, e.g. what Julien means is:
>> -#define XENDISPL_EVENT_PAGE_SIZE 4096
>> +#define XENDISPL_EVENT_PAGE_SIZE XEN_PAGE_SIZE
>> Julien, does this change work for you?
> Sure, I think that would be OK. Also using XEN_PAGE_SIZE directly would
> be OK.
>
Great, thanks
>>>>>> +#define XENDISPL_IN_RING_OFFS (sizeof(struct xendispl_event_page))
>>>>>> +#define XENDISPL_IN_RING_SIZE (XENDISPL_EVENT_PAGE_SIZE -
>>>>>> XENDISPL_IN_RING_OFFS)
>>>>>> +#define XENDISPL_IN_RING_LEN (XENDISPL_IN_RING_SIZE / sizeof(struct
>>>>>> xendispl_evt))
>>>>>> +#define XENDISPL_IN_RING(page) \
>>>>>> + ((struct xendispl_evt *)((char *)(page) + XENDISPL_IN_RING_OFFS))
>>>>>> +#define XENDISPL_IN_RING_REF(page, idx) \
>>>>>> + (XENDISPL_IN_RING((page))[(idx) % XENDISPL_IN_RING_LEN])
>>>>>> +
>>>>>> +#endif /* __XEN_PUBLIC_IO_DISPLIF_H__ */
>>>>>> +
>>>>>> +/*
>>>>>> + * Local variables:
>>>>>> + * mode: C
>>>>>> + * c-file-style: "BSD"
>>>>>> + * c-basic-offset: 4
>>>>>> + * tab-width: 4
>>>>>> + * indent-tabs-mode: nil
>>>>>> + * End:
>>>>>> + */
>>>>>>
>>>>> Cheers,
>>>>>
>>>> Thank you,
>>>> Oleksandr
>>>>
>> Thank you,
>> Oleksandr
>>
_______________________________________________
Xen-devel mailing list
Xen-devel@lists.xen.org
https://lists.xen.org/xen-devel
^ permalink raw reply [flat|nested] 41+ messages in thread
* Re: [For Linux 4/4] xen/displif: add ABI for para-virtual display
2017-04-07 14:02 ` Oleksandr Andrushchenko
` (2 preceding siblings ...)
2017-04-10 6:03 ` Juergen Gross
@ 2017-04-10 6:03 ` Juergen Gross
2017-04-10 6:06 ` Oleksandr Andrushchenko
2017-04-10 6:06 ` Oleksandr Andrushchenko
3 siblings, 2 replies; 41+ messages in thread
From: Juergen Gross @ 2017-04-10 6:03 UTC (permalink / raw)
To: Oleksandr Andrushchenko, Julien Grall, xen-devel
Cc: joculator, al1img, vlad.babchuk, andrii.anisov, olekstysh,
konrad.wilk, lars.kurth, sstabellini, linux-kernel,
Oleksandr Andrushchenko, Oleksandr Grytsov
On 07/04/17 16:02, Oleksandr Andrushchenko wrote:
> Hi, Julien!
>
> On 04/07/2017 04:50 PM, Julien Grall wrote:
>> Hi Oleksandr,
>>
>> On 07/04/17 09:30, Oleksandr Andrushchenko wrote:
>>> +/*
>>> +
>>> ******************************************************************************
>>>
>>> + * Back to front events delivery
>>> +
>>> ******************************************************************************
>>>
>>> + * In order to deliver asynchronous events from back to front a
>>> shared page is
>>> + * allocated by front and its granted reference propagated to back via
>>> + * XenStore entries (evt-ring-ref/evt-event-channel).
>>> + * This page has a common header used by both front and back to
>>> synchronize
>>> + * access and control event's ring buffer, while back being a
>>> producer of the
>>> + * events and front being a consumer. The rest of the page after the
>>> header
>>> + * is used for event packets.
>>> + *
>>> + * Upon reception of an event(s) front may confirm its reception
>>> + * for either each event, group of events or none.
>>> + */
>>> +
>>> +struct xendispl_event_page {
>>> + uint32_t in_cons;
>>> + uint32_t in_prod;
>>> + uint8_t reserved[56];
>>> +};
>>> +
>>> +#define XENDISPL_EVENT_PAGE_SIZE 4096
>>
>> This will be always the size of a grant (e.g xen page size), right? If
>> so, I would prefer if you use XEN_PAGE_SIZE to so we can easily update
>> Linux in the case Xen is using a different page size.
>>
>> Note that the Front-end, Backend and Xen may use different page size,
>> so your drivers would need to cope with that. The current approach is
>> to always use the page granularity of Xen.
>>
> Will the following satisfy the requirement?
> #define XENDISPL_EVENT_PAGE_SIZE XEN_PAGE_SIZE
>
>>> +#define XENDISPL_IN_RING_OFFS (sizeof(struct xendispl_event_page))
>>> +#define XENDISPL_IN_RING_SIZE (XENDISPL_EVENT_PAGE_SIZE -
>>> XENDISPL_IN_RING_OFFS)
>>> +#define XENDISPL_IN_RING_LEN (XENDISPL_IN_RING_SIZE / sizeof(struct
>>> xendispl_evt))
>>> +#define XENDISPL_IN_RING(page) \
>>> + ((struct xendispl_evt *)((char *)(page) + XENDISPL_IN_RING_OFFS))
>>> +#define XENDISPL_IN_RING_REF(page, idx) \
>>> + (XENDISPL_IN_RING((page))[(idx) % XENDISPL_IN_RING_LEN])
>>> +
>>> +#endif /* __XEN_PUBLIC_IO_DISPLIF_H__ */
>>> +
>>> +/*
>>> + * Local variables:
>>> + * mode: C
>>> + * c-file-style: "BSD"
>>> + * c-basic-offset: 4
>>> + * tab-width: 4
>>> + * indent-tabs-mode: nil
>>> + * End:
>>> + */
I believe you'll send V2 due to the XEN_PAGE_SIZE usage?
Can you then please remove the editor mode settings block in each
header, too? Those are disliked in the Linux tree.
Juergen
^ permalink raw reply [flat|nested] 41+ messages in thread
* Re: [For Linux 4/4] xen/displif: add ABI for para-virtual display
2017-04-07 14:02 ` Oleksandr Andrushchenko
2017-04-07 16:36 ` Stefano Stabellini
2017-04-07 16:36 ` Stefano Stabellini
@ 2017-04-10 6:03 ` Juergen Gross
2017-04-10 6:03 ` Juergen Gross
3 siblings, 0 replies; 41+ messages in thread
From: Juergen Gross @ 2017-04-10 6:03 UTC (permalink / raw)
To: Oleksandr Andrushchenko, Julien Grall, xen-devel
Cc: lars.kurth, sstabellini, vlad.babchuk, Oleksandr Andrushchenko,
linux-kernel, andrii.anisov, olekstysh, al1img,
Oleksandr Grytsov, joculator
On 07/04/17 16:02, Oleksandr Andrushchenko wrote:
> Hi, Julien!
>
> On 04/07/2017 04:50 PM, Julien Grall wrote:
>> Hi Oleksandr,
>>
>> On 07/04/17 09:30, Oleksandr Andrushchenko wrote:
>>> +/*
>>> +
>>> ******************************************************************************
>>>
>>> + * Back to front events delivery
>>> +
>>> ******************************************************************************
>>>
>>> + * In order to deliver asynchronous events from back to front a
>>> shared page is
>>> + * allocated by front and its granted reference propagated to back via
>>> + * XenStore entries (evt-ring-ref/evt-event-channel).
>>> + * This page has a common header used by both front and back to
>>> synchronize
>>> + * access and control event's ring buffer, while back being a
>>> producer of the
>>> + * events and front being a consumer. The rest of the page after the
>>> header
>>> + * is used for event packets.
>>> + *
>>> + * Upon reception of an event(s) front may confirm its reception
>>> + * for either each event, group of events or none.
>>> + */
>>> +
>>> +struct xendispl_event_page {
>>> + uint32_t in_cons;
>>> + uint32_t in_prod;
>>> + uint8_t reserved[56];
>>> +};
>>> +
>>> +#define XENDISPL_EVENT_PAGE_SIZE 4096
>>
>> This will be always the size of a grant (e.g xen page size), right? If
>> so, I would prefer if you use XEN_PAGE_SIZE to so we can easily update
>> Linux in the case Xen is using a different page size.
>>
>> Note that the Front-end, Backend and Xen may use different page size,
>> so your drivers would need to cope with that. The current approach is
>> to always use the page granularity of Xen.
>>
> Will the following satisfy the requirement?
> #define XENDISPL_EVENT_PAGE_SIZE XEN_PAGE_SIZE
>
>>> +#define XENDISPL_IN_RING_OFFS (sizeof(struct xendispl_event_page))
>>> +#define XENDISPL_IN_RING_SIZE (XENDISPL_EVENT_PAGE_SIZE -
>>> XENDISPL_IN_RING_OFFS)
>>> +#define XENDISPL_IN_RING_LEN (XENDISPL_IN_RING_SIZE / sizeof(struct
>>> xendispl_evt))
>>> +#define XENDISPL_IN_RING(page) \
>>> + ((struct xendispl_evt *)((char *)(page) + XENDISPL_IN_RING_OFFS))
>>> +#define XENDISPL_IN_RING_REF(page, idx) \
>>> + (XENDISPL_IN_RING((page))[(idx) % XENDISPL_IN_RING_LEN])
>>> +
>>> +#endif /* __XEN_PUBLIC_IO_DISPLIF_H__ */
>>> +
>>> +/*
>>> + * Local variables:
>>> + * mode: C
>>> + * c-file-style: "BSD"
>>> + * c-basic-offset: 4
>>> + * tab-width: 4
>>> + * indent-tabs-mode: nil
>>> + * End:
>>> + */
I believe you'll send V2 due to the XEN_PAGE_SIZE usage?
Can you then please remove the editor mode settings block in each
header, too? Those are disliked in the Linux tree.
Juergen
_______________________________________________
Xen-devel mailing list
Xen-devel@lists.xen.org
https://lists.xen.org/xen-devel
^ permalink raw reply [flat|nested] 41+ messages in thread
* Re: [For Linux 4/4] xen/displif: add ABI for para-virtual display
2017-04-10 6:03 ` Juergen Gross
@ 2017-04-10 6:06 ` Oleksandr Andrushchenko
2017-04-10 6:07 ` Juergen Gross
2017-04-10 6:07 ` Juergen Gross
2017-04-10 6:06 ` Oleksandr Andrushchenko
1 sibling, 2 replies; 41+ messages in thread
From: Oleksandr Andrushchenko @ 2017-04-10 6:06 UTC (permalink / raw)
To: Juergen Gross, Julien Grall, xen-devel
Cc: joculator, al1img, vlad.babchuk, andrii.anisov, olekstysh,
konrad.wilk, lars.kurth, sstabellini, linux-kernel,
Oleksandr Andrushchenko, Oleksandr Grytsov
On 04/10/2017 09:03 AM, Juergen Gross wrote:
> On 07/04/17 16:02, Oleksandr Andrushchenko wrote:
>> Hi, Julien!
>>
>> On 04/07/2017 04:50 PM, Julien Grall wrote:
>>> Hi Oleksandr,
>>>
>>> On 07/04/17 09:30, Oleksandr Andrushchenko wrote:
>>>> +/*
>>>> +
>>>> ******************************************************************************
>>>>
>>>> + * Back to front events delivery
>>>> +
>>>> ******************************************************************************
>>>>
>>>> + * In order to deliver asynchronous events from back to front a
>>>> shared page is
>>>> + * allocated by front and its granted reference propagated to back via
>>>> + * XenStore entries (evt-ring-ref/evt-event-channel).
>>>> + * This page has a common header used by both front and back to
>>>> synchronize
>>>> + * access and control event's ring buffer, while back being a
>>>> producer of the
>>>> + * events and front being a consumer. The rest of the page after the
>>>> header
>>>> + * is used for event packets.
>>>> + *
>>>> + * Upon reception of an event(s) front may confirm its reception
>>>> + * for either each event, group of events or none.
>>>> + */
>>>> +
>>>> +struct xendispl_event_page {
>>>> + uint32_t in_cons;
>>>> + uint32_t in_prod;
>>>> + uint8_t reserved[56];
>>>> +};
>>>> +
>>>> +#define XENDISPL_EVENT_PAGE_SIZE 4096
>>> This will be always the size of a grant (e.g xen page size), right? If
>>> so, I would prefer if you use XEN_PAGE_SIZE to so we can easily update
>>> Linux in the case Xen is using a different page size.
>>>
>>> Note that the Front-end, Backend and Xen may use different page size,
>>> so your drivers would need to cope with that. The current approach is
>>> to always use the page granularity of Xen.
>>>
>> Will the following satisfy the requirement?
>> #define XENDISPL_EVENT_PAGE_SIZE XEN_PAGE_SIZE
>>
>>>> +#define XENDISPL_IN_RING_OFFS (sizeof(struct xendispl_event_page))
>>>> +#define XENDISPL_IN_RING_SIZE (XENDISPL_EVENT_PAGE_SIZE -
>>>> XENDISPL_IN_RING_OFFS)
>>>> +#define XENDISPL_IN_RING_LEN (XENDISPL_IN_RING_SIZE / sizeof(struct
>>>> xendispl_evt))
>>>> +#define XENDISPL_IN_RING(page) \
>>>> + ((struct xendispl_evt *)((char *)(page) + XENDISPL_IN_RING_OFFS))
>>>> +#define XENDISPL_IN_RING_REF(page, idx) \
>>>> + (XENDISPL_IN_RING((page))[(idx) % XENDISPL_IN_RING_LEN])
>>>> +
>>>> +#endif /* __XEN_PUBLIC_IO_DISPLIF_H__ */
>>>> +
>>>> +/*
>>>> + * Local variables:
>>>> + * mode: C
>>>> + * c-file-style: "BSD"
>>>> + * c-basic-offset: 4
>>>> + * tab-width: 4
>>>> + * indent-tabs-mode: nil
>>>> + * End:
>>>> + */
> I believe you'll send V2 due to the XEN_PAGE_SIZE usage?
yes
>
> Can you then please remove the editor mode settings block in each
> header, too? Those are disliked in the Linux tree.
sure, I will
>
>
> Juergen
Do you mind if I keep your acks for V2 with these changes?
Thank you,
Oleksandr
^ permalink raw reply [flat|nested] 41+ messages in thread
* Re: [For Linux 4/4] xen/displif: add ABI for para-virtual display
2017-04-10 6:03 ` Juergen Gross
2017-04-10 6:06 ` Oleksandr Andrushchenko
@ 2017-04-10 6:06 ` Oleksandr Andrushchenko
1 sibling, 0 replies; 41+ messages in thread
From: Oleksandr Andrushchenko @ 2017-04-10 6:06 UTC (permalink / raw)
To: Juergen Gross, Julien Grall, xen-devel
Cc: lars.kurth, sstabellini, vlad.babchuk, Oleksandr Andrushchenko,
linux-kernel, andrii.anisov, olekstysh, al1img,
Oleksandr Grytsov, joculator
On 04/10/2017 09:03 AM, Juergen Gross wrote:
> On 07/04/17 16:02, Oleksandr Andrushchenko wrote:
>> Hi, Julien!
>>
>> On 04/07/2017 04:50 PM, Julien Grall wrote:
>>> Hi Oleksandr,
>>>
>>> On 07/04/17 09:30, Oleksandr Andrushchenko wrote:
>>>> +/*
>>>> +
>>>> ******************************************************************************
>>>>
>>>> + * Back to front events delivery
>>>> +
>>>> ******************************************************************************
>>>>
>>>> + * In order to deliver asynchronous events from back to front a
>>>> shared page is
>>>> + * allocated by front and its granted reference propagated to back via
>>>> + * XenStore entries (evt-ring-ref/evt-event-channel).
>>>> + * This page has a common header used by both front and back to
>>>> synchronize
>>>> + * access and control event's ring buffer, while back being a
>>>> producer of the
>>>> + * events and front being a consumer. The rest of the page after the
>>>> header
>>>> + * is used for event packets.
>>>> + *
>>>> + * Upon reception of an event(s) front may confirm its reception
>>>> + * for either each event, group of events or none.
>>>> + */
>>>> +
>>>> +struct xendispl_event_page {
>>>> + uint32_t in_cons;
>>>> + uint32_t in_prod;
>>>> + uint8_t reserved[56];
>>>> +};
>>>> +
>>>> +#define XENDISPL_EVENT_PAGE_SIZE 4096
>>> This will be always the size of a grant (e.g xen page size), right? If
>>> so, I would prefer if you use XEN_PAGE_SIZE to so we can easily update
>>> Linux in the case Xen is using a different page size.
>>>
>>> Note that the Front-end, Backend and Xen may use different page size,
>>> so your drivers would need to cope with that. The current approach is
>>> to always use the page granularity of Xen.
>>>
>> Will the following satisfy the requirement?
>> #define XENDISPL_EVENT_PAGE_SIZE XEN_PAGE_SIZE
>>
>>>> +#define XENDISPL_IN_RING_OFFS (sizeof(struct xendispl_event_page))
>>>> +#define XENDISPL_IN_RING_SIZE (XENDISPL_EVENT_PAGE_SIZE -
>>>> XENDISPL_IN_RING_OFFS)
>>>> +#define XENDISPL_IN_RING_LEN (XENDISPL_IN_RING_SIZE / sizeof(struct
>>>> xendispl_evt))
>>>> +#define XENDISPL_IN_RING(page) \
>>>> + ((struct xendispl_evt *)((char *)(page) + XENDISPL_IN_RING_OFFS))
>>>> +#define XENDISPL_IN_RING_REF(page, idx) \
>>>> + (XENDISPL_IN_RING((page))[(idx) % XENDISPL_IN_RING_LEN])
>>>> +
>>>> +#endif /* __XEN_PUBLIC_IO_DISPLIF_H__ */
>>>> +
>>>> +/*
>>>> + * Local variables:
>>>> + * mode: C
>>>> + * c-file-style: "BSD"
>>>> + * c-basic-offset: 4
>>>> + * tab-width: 4
>>>> + * indent-tabs-mode: nil
>>>> + * End:
>>>> + */
> I believe you'll send V2 due to the XEN_PAGE_SIZE usage?
yes
>
> Can you then please remove the editor mode settings block in each
> header, too? Those are disliked in the Linux tree.
sure, I will
>
>
> Juergen
Do you mind if I keep your acks for V2 with these changes?
Thank you,
Oleksandr
_______________________________________________
Xen-devel mailing list
Xen-devel@lists.xen.org
https://lists.xen.org/xen-devel
^ permalink raw reply [flat|nested] 41+ messages in thread
* Re: [For Linux 4/4] xen/displif: add ABI for para-virtual display
2017-04-10 6:06 ` Oleksandr Andrushchenko
@ 2017-04-10 6:07 ` Juergen Gross
2017-04-10 6:07 ` Juergen Gross
1 sibling, 0 replies; 41+ messages in thread
From: Juergen Gross @ 2017-04-10 6:07 UTC (permalink / raw)
To: Oleksandr Andrushchenko, Julien Grall, xen-devel
Cc: joculator, al1img, vlad.babchuk, andrii.anisov, olekstysh,
konrad.wilk, lars.kurth, sstabellini, linux-kernel,
Oleksandr Andrushchenko, Oleksandr Grytsov
On 10/04/17 08:06, Oleksandr Andrushchenko wrote:
> On 04/10/2017 09:03 AM, Juergen Gross wrote:
>> On 07/04/17 16:02, Oleksandr Andrushchenko wrote:
>>> Hi, Julien!
>>>
>>> On 04/07/2017 04:50 PM, Julien Grall wrote:
>>>> Hi Oleksandr,
>>>>
>>>> On 07/04/17 09:30, Oleksandr Andrushchenko wrote:
>>>>> +/*
>>>>> +
>>>>> ******************************************************************************
>>>>>
>>>>>
>>>>> + * Back to front events delivery
>>>>> +
>>>>> ******************************************************************************
>>>>>
>>>>>
>>>>> + * In order to deliver asynchronous events from back to front a
>>>>> shared page is
>>>>> + * allocated by front and its granted reference propagated to back
>>>>> via
>>>>> + * XenStore entries (evt-ring-ref/evt-event-channel).
>>>>> + * This page has a common header used by both front and back to
>>>>> synchronize
>>>>> + * access and control event's ring buffer, while back being a
>>>>> producer of the
>>>>> + * events and front being a consumer. The rest of the page after the
>>>>> header
>>>>> + * is used for event packets.
>>>>> + *
>>>>> + * Upon reception of an event(s) front may confirm its reception
>>>>> + * for either each event, group of events or none.
>>>>> + */
>>>>> +
>>>>> +struct xendispl_event_page {
>>>>> + uint32_t in_cons;
>>>>> + uint32_t in_prod;
>>>>> + uint8_t reserved[56];
>>>>> +};
>>>>> +
>>>>> +#define XENDISPL_EVENT_PAGE_SIZE 4096
>>>> This will be always the size of a grant (e.g xen page size), right? If
>>>> so, I would prefer if you use XEN_PAGE_SIZE to so we can easily update
>>>> Linux in the case Xen is using a different page size.
>>>>
>>>> Note that the Front-end, Backend and Xen may use different page size,
>>>> so your drivers would need to cope with that. The current approach is
>>>> to always use the page granularity of Xen.
>>>>
>>> Will the following satisfy the requirement?
>>> #define XENDISPL_EVENT_PAGE_SIZE XEN_PAGE_SIZE
>>>
>>>>> +#define XENDISPL_IN_RING_OFFS (sizeof(struct xendispl_event_page))
>>>>> +#define XENDISPL_IN_RING_SIZE (XENDISPL_EVENT_PAGE_SIZE -
>>>>> XENDISPL_IN_RING_OFFS)
>>>>> +#define XENDISPL_IN_RING_LEN (XENDISPL_IN_RING_SIZE / sizeof(struct
>>>>> xendispl_evt))
>>>>> +#define XENDISPL_IN_RING(page) \
>>>>> + ((struct xendispl_evt *)((char *)(page) + XENDISPL_IN_RING_OFFS))
>>>>> +#define XENDISPL_IN_RING_REF(page, idx) \
>>>>> + (XENDISPL_IN_RING((page))[(idx) % XENDISPL_IN_RING_LEN])
>>>>> +
>>>>> +#endif /* __XEN_PUBLIC_IO_DISPLIF_H__ */
>>>>> +
>>>>> +/*
>>>>> + * Local variables:
>>>>> + * mode: C
>>>>> + * c-file-style: "BSD"
>>>>> + * c-basic-offset: 4
>>>>> + * tab-width: 4
>>>>> + * indent-tabs-mode: nil
>>>>> + * End:
>>>>> + */
>> I believe you'll send V2 due to the XEN_PAGE_SIZE usage?
> yes
>>
>> Can you then please remove the editor mode settings block in each
>> header, too? Those are disliked in the Linux tree.
> sure, I will
>>
>>
>> Juergen
> Do you mind if I keep your acks for V2 with these changes?
You can keep them, of course.
Juergen
^ permalink raw reply [flat|nested] 41+ messages in thread
* Re: [For Linux 4/4] xen/displif: add ABI for para-virtual display
2017-04-10 6:06 ` Oleksandr Andrushchenko
2017-04-10 6:07 ` Juergen Gross
@ 2017-04-10 6:07 ` Juergen Gross
1 sibling, 0 replies; 41+ messages in thread
From: Juergen Gross @ 2017-04-10 6:07 UTC (permalink / raw)
To: Oleksandr Andrushchenko, Julien Grall, xen-devel
Cc: lars.kurth, sstabellini, vlad.babchuk, Oleksandr Andrushchenko,
linux-kernel, andrii.anisov, olekstysh, al1img,
Oleksandr Grytsov, joculator
On 10/04/17 08:06, Oleksandr Andrushchenko wrote:
> On 04/10/2017 09:03 AM, Juergen Gross wrote:
>> On 07/04/17 16:02, Oleksandr Andrushchenko wrote:
>>> Hi, Julien!
>>>
>>> On 04/07/2017 04:50 PM, Julien Grall wrote:
>>>> Hi Oleksandr,
>>>>
>>>> On 07/04/17 09:30, Oleksandr Andrushchenko wrote:
>>>>> +/*
>>>>> +
>>>>> ******************************************************************************
>>>>>
>>>>>
>>>>> + * Back to front events delivery
>>>>> +
>>>>> ******************************************************************************
>>>>>
>>>>>
>>>>> + * In order to deliver asynchronous events from back to front a
>>>>> shared page is
>>>>> + * allocated by front and its granted reference propagated to back
>>>>> via
>>>>> + * XenStore entries (evt-ring-ref/evt-event-channel).
>>>>> + * This page has a common header used by both front and back to
>>>>> synchronize
>>>>> + * access and control event's ring buffer, while back being a
>>>>> producer of the
>>>>> + * events and front being a consumer. The rest of the page after the
>>>>> header
>>>>> + * is used for event packets.
>>>>> + *
>>>>> + * Upon reception of an event(s) front may confirm its reception
>>>>> + * for either each event, group of events or none.
>>>>> + */
>>>>> +
>>>>> +struct xendispl_event_page {
>>>>> + uint32_t in_cons;
>>>>> + uint32_t in_prod;
>>>>> + uint8_t reserved[56];
>>>>> +};
>>>>> +
>>>>> +#define XENDISPL_EVENT_PAGE_SIZE 4096
>>>> This will be always the size of a grant (e.g xen page size), right? If
>>>> so, I would prefer if you use XEN_PAGE_SIZE to so we can easily update
>>>> Linux in the case Xen is using a different page size.
>>>>
>>>> Note that the Front-end, Backend and Xen may use different page size,
>>>> so your drivers would need to cope with that. The current approach is
>>>> to always use the page granularity of Xen.
>>>>
>>> Will the following satisfy the requirement?
>>> #define XENDISPL_EVENT_PAGE_SIZE XEN_PAGE_SIZE
>>>
>>>>> +#define XENDISPL_IN_RING_OFFS (sizeof(struct xendispl_event_page))
>>>>> +#define XENDISPL_IN_RING_SIZE (XENDISPL_EVENT_PAGE_SIZE -
>>>>> XENDISPL_IN_RING_OFFS)
>>>>> +#define XENDISPL_IN_RING_LEN (XENDISPL_IN_RING_SIZE / sizeof(struct
>>>>> xendispl_evt))
>>>>> +#define XENDISPL_IN_RING(page) \
>>>>> + ((struct xendispl_evt *)((char *)(page) + XENDISPL_IN_RING_OFFS))
>>>>> +#define XENDISPL_IN_RING_REF(page, idx) \
>>>>> + (XENDISPL_IN_RING((page))[(idx) % XENDISPL_IN_RING_LEN])
>>>>> +
>>>>> +#endif /* __XEN_PUBLIC_IO_DISPLIF_H__ */
>>>>> +
>>>>> +/*
>>>>> + * Local variables:
>>>>> + * mode: C
>>>>> + * c-file-style: "BSD"
>>>>> + * c-basic-offset: 4
>>>>> + * tab-width: 4
>>>>> + * indent-tabs-mode: nil
>>>>> + * End:
>>>>> + */
>> I believe you'll send V2 due to the XEN_PAGE_SIZE usage?
> yes
>>
>> Can you then please remove the editor mode settings block in each
>> header, too? Those are disliked in the Linux tree.
> sure, I will
>>
>>
>> Juergen
> Do you mind if I keep your acks for V2 with these changes?
You can keep them, of course.
Juergen
_______________________________________________
Xen-devel mailing list
Xen-devel@lists.xen.org
https://lists.xen.org/xen-devel
^ permalink raw reply [flat|nested] 41+ messages in thread
end of thread, other threads:[~2017-04-10 6:07 UTC | newest]
Thread overview: 41+ messages (download: mbox.gz / follow: Atom feed)
-- links below jump to the message on this page --
2017-04-07 8:30 [For Linux 0/4] PV protocol headers for Linux Kernel Oleksandr Andrushchenko
2017-04-07 8:30 ` [For Linux 1/4] xen/kbdif: update protocol description Oleksandr Andrushchenko
2017-04-07 8:30 ` Oleksandr Andrushchenko
2017-04-07 11:29 ` [Xen-devel] " Juergen Gross
2017-04-07 12:45 ` Konrad Rzeszutek Wilk
2017-04-07 12:45 ` [Xen-devel] " Konrad Rzeszutek Wilk
2017-04-07 12:58 ` Oleksandr Andrushchenko
2017-04-07 12:58 ` [Xen-devel] " Oleksandr Andrushchenko
2017-04-07 11:29 ` Juergen Gross
2017-04-07 8:30 ` [For Linux 2/4] xen/kbdif: add multi-touch support Oleksandr Andrushchenko
2017-04-07 8:30 ` Oleksandr Andrushchenko
2017-04-07 11:30 ` Juergen Gross
2017-04-07 11:30 ` [Xen-devel] " Juergen Gross
2017-04-07 8:30 ` [For Linux 3/4] xen/sndif: add sound-device ABI Oleksandr Andrushchenko
2017-04-07 8:30 ` Oleksandr Andrushchenko
2017-04-07 11:31 ` Juergen Gross
2017-04-07 11:31 ` [Xen-devel] " Juergen Gross
2017-04-07 8:30 ` [For Linux 4/4] xen/displif: add ABI for para-virtual display Oleksandr Andrushchenko
2017-04-07 11:31 ` [Xen-devel] " Juergen Gross
2017-04-07 11:31 ` Juergen Gross
2017-04-07 13:50 ` Julien Grall
2017-04-07 14:02 ` Oleksandr Andrushchenko
2017-04-07 14:02 ` Oleksandr Andrushchenko
2017-04-07 16:36 ` Stefano Stabellini
2017-04-07 17:43 ` Oleksandr Andrushchenko
2017-04-07 17:43 ` Oleksandr Andrushchenko
2017-04-07 17:50 ` Stefano Stabellini
2017-04-07 17:50 ` Stefano Stabellini
2017-04-07 17:51 ` Oleksandr Andrushchenko
2017-04-07 17:51 ` Oleksandr Andrushchenko
2017-04-07 16:36 ` Stefano Stabellini
2017-04-10 6:03 ` Juergen Gross
2017-04-10 6:03 ` Juergen Gross
2017-04-10 6:06 ` Oleksandr Andrushchenko
2017-04-10 6:07 ` Juergen Gross
2017-04-10 6:07 ` Juergen Gross
2017-04-10 6:06 ` Oleksandr Andrushchenko
2017-04-07 13:50 ` Julien Grall
2017-04-07 8:30 ` Oleksandr Andrushchenko
2017-04-07 8:47 ` [For Linux 0/4] PV protocol headers for Linux Kernel Oleksandr Andrushchenko
2017-04-07 8:47 ` Oleksandr Andrushchenko
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.