[PATCH v5 0/4] new ring macros, 9pfs and pvcalls headers

[PATCH v5 0/4] new ring macros, 9pfs and pvcalls headers

[Stefano Stabellini]

Hi all,

this patch series introduces a set of new ring macros to support rings
in the formats specified by the Xen 9pfs transport and PV Calls
protocol. It also introduces the Xen 9pfs and PV Calls protocols
headers.

Changes in v5:
- parenthesize uses of macro parameters in XEN_FLEX_RING_SIZE
- add grant_table.h to the list of prereqs in ring.h and remove the
  #include from ring.h
- #include grant_table.h in 9pfs.h and pvcalls.h
- remove PAGE_SHIFT definition, define XEN_PAGE_SHIFT instead
- code style fixes
- remove struct xen_9pfs_header definition
- don't add extra -include cstring to all c++ tests, only when needed
- add headers99.chk to .gitignore 

Changes in v4:
- include ../grant_table.h in ring.h
- add a comment about required declarations on top of ring.h
- add a patch to introduce a C99 headers check
- add -include string.h to the existing C++ headers check
- add 9pfs and pvcalls to the C99 headers check

Changes in v3:
- fix commit message
- add newlines after read/write_packet functions
- reorder DEFINE_XEN_FLEX_RING_AND_INTF and DEFINE_XEN_FLEX_RING

Changes in v2:
- replace __attribute__((packed)) with #pragma pack
- remove XEN_9PFS_RING_ORDER, the 9pfs ring order should be dynamic
- add editor configuration blocks
- add links to the specs


Stefano Stabellini (4):
      ring.h: introduce macros to handle monodirectional rings with multiple req sizes
      xen: introduce a C99 headers check
      Introduce the Xen 9pfs transport header
      Introduce the pvcalls header

 .gitignore                      |   1 +
 xen/include/Makefile            |  29 ++++++--
 xen/include/public/io/9pfs.h    |  49 +++++++++++++
 xen/include/public/io/pvcalls.h | 153 ++++++++++++++++++++++++++++++++++++++++
 xen/include/public/io/ring.h    | 150 +++++++++++++++++++++++++++++++++++++++
 5 files changed, 378 insertions(+), 4 deletions(-)
 create mode 100644 xen/include/public/io/9pfs.h
 create mode 100644 xen/include/public/io/pvcalls.h

Cheers,

Stefano

_______________________________________________
Xen-devel mailing list
Xen-devel@lists.xen.org
https://lists.xen.org/xen-devel

[PATCH v5 1/4] ring.h: introduce macros to handle monodirectional rings with multiple req sizes

[Stefano Stabellini]

This patch introduces macros, structs and functions to handle rings in
the format described by docs/misc/pvcalls.markdown and
docs/misc/9pfs.markdown. The index page (struct __name##_data_intf)
contains the indexes and the grant refs to setup two rings.

               Indexes page
               +----------------------+
               |@0 $NAME_data_intf:   |
               |@76: ring_order = 1   |
               |@80: ref[0]+          |
               |@84: ref[1]+          |
               |           |          |
               |           |          |
               +----------------------+
                           |
                           v (data ring)
                   +-------+-----------+
                   |  @0->4095: in     |
                   |  ref[0]           |
                   |-------------------|
                   |  @4096->8191: out |
                   |  ref[1]           |
                   +-------------------+

$NAME_read_packet and $NAME_write_packet are provided to read or write
any data struct from/to the ring. In pvcalls, they are unused. In xen
9pfs, they are used to read or write the 9pfs header. In other protocols
they could be used to read/write the whole request structure. See
docs/misc/9pfs.markdown:Ring Usage to learn how to check how much data
is on the ring, and how to handle notifications.

There is a ring_size parameter to most functions so that protocols using
these macros don't have to have a statically defined ring order at build
time. In pvcalls for example, each new ring could have a different
order.

These macros don't help you share the indexes page or the event channels
needed for notifications. You can do that with other out of band
mechanisms, such as xenstore or another ring.

It is not possible to use a macro to define another macro with a
variable name. For this reason, this patch introduces static inline
functions instead, that are not C89 compliant. Additionally, the macro
defines a struct with a variable sized array, which is also not C89
compliant.

Signed-off-by: Stefano Stabellini <stefano@aporeto.com>
CC: konrad.wilk@oracle.com
CC: JBeulich@suse.com
---
 xen/include/public/io/ring.h | 150 +++++++++++++++++++++++++++++++++++++++++++
 1 file changed, 150 insertions(+)

diff --git a/xen/include/public/io/ring.h b/xen/include/public/io/ring.h
index 801c0da..89c48d7 100644
--- a/xen/include/public/io/ring.h
+++ b/xen/include/public/io/ring.h
@@ -27,6 +27,21 @@
 #ifndef __XEN_PUBLIC_IO_RING_H__
 #define __XEN_PUBLIC_IO_RING_H__
 
+/*
+ * When #include'ing this header, you need to provide the following
+ * declaration upfront:
+ * - standard integers types (uint8_t, uint16_t, etc)
+ * They are provided by stdint.h of the standard headers.
+ *
+ * In addition, if you intend to use the FLEX macros, you also need to
+ * provide:
+ * - size_t
+ * - memcpy
+ * - grant_ref_t
+ * These declarations are provided by string.h of the standard headers,
+ * and grant_table.h from the Xen public headers.
+ */
+
 #include "../xen-compat.h"
 
 #if __XEN_INTERFACE_VERSION__ < 0x00030208
@@ -313,6 +328,141 @@ typedef struct __name##_back_ring __name##_back_ring_t
     (_work_to_do) = RING_HAS_UNCONSUMED_RESPONSES(_r);                  \
 } while (0)
 
+
+/*
+ * DEFINE_XEN_FLEX_RING_AND_INTF defines two monodirectional rings and
+ * functions to check if there is data on the ring, and to read and
+ * write to them.
+ *
+ * DEFINE_XEN_FLEX_RING is similar to DEFINE_XEN_FLEX_RING_AND_INTF, but
+ * does not define the indexes page. As different protocols can have
+ * extensions to the basic format, this macro allow them to define their
+ * own struct.
+ *
+ * XEN_FLEX_RING_SIZE
+ *   Convenience macro to calculate the size of one of the two rings
+ *   from the overall order.
+ *
+ * $NAME_mask
+ *   Function to apply the size mask to an index, to reduce the index
+ *   within the range [0-size].
+ *
+ * $NAME_read_packet
+ *   Function to read data from the ring. The amount of data to read is
+ *   specified by the "size" argument.
+ *
+ * $NAME_write_packet
+ *   Function to write data to the ring. The amount of data to write is
+ *   specified by the "size" argument.
+ *
+ * $NAME_get_ring_ptr
+ *   Convenience function that returns a pointer to read/write to the
+ *   ring at the right location.
+ *
+ * $NAME_data_intf
+ *   Indexes page, shared between frontend and backend. It also
+ *   contains the array of grant refs.
+ *
+ * $NAME_queued
+ *   Function to calculate how many bytes are currently on the ring,
+ *   ready to be read. It can also be used to calculate how much free
+ *   space is currently on the ring (ring_size - $NAME_queued()).
+ */
+
+#define XEN_PAGE_SHIFT 12
+#define XEN_FLEX_RING_SIZE(order)                                             \
+    (1UL << ((order) + XEN_PAGE_SHIFT - 1))
+
+#define DEFINE_XEN_FLEX_RING(name)                                            \
+static inline RING_IDX name##_mask(RING_IDX idx, RING_IDX ring_size)          \
+{                                                                             \
+    return (idx & (ring_size - 1));                                           \
+}                                                                             \
+                                                                              \
+static inline RING_IDX name##_mask_order(RING_IDX idx, RING_IDX ring_order)   \
+{                                                                             \
+    return (idx & (XEN_FLEX_RING_SIZE(ring_order) - 1));                      \
+}                                                                             \
+                                                                              \
+static inline unsigned char *name##_get_ring_ptr(unsigned char *buf,          \
+                                                 RING_IDX idx,                \
+                                                 RING_IDX ring_order)         \
+{                                                                             \
+    return buf + name##_mask_order(idx, ring_order);                          \
+}                                                                             \
+                                                                              \
+static inline void name##_read_packet(const unsigned char *buf,               \
+        RING_IDX masked_prod, RING_IDX *masked_cons,                          \
+        RING_IDX ring_size, void *opaque, size_t size)                        \
+{                                                                             \
+    if (*masked_cons < masked_prod ||                                         \
+            size <= ring_size - *masked_cons) {                               \
+        memcpy(opaque, buf + *masked_cons, size);                             \
+    } else {                                                                  \
+        memcpy(opaque, buf + *masked_cons, ring_size - *masked_cons);         \
+        memcpy((unsigned char *)opaque + ring_size - *masked_cons, buf,       \
+                size - (ring_size - *masked_cons));                           \
+    }                                                                         \
+    *masked_cons = name##_mask(*masked_cons + size, ring_size);               \
+}                                                                             \
+                                                                              \
+static inline void name##_write_packet(unsigned char *buf,                    \
+        RING_IDX *masked_prod, RING_IDX masked_cons,                          \
+        RING_IDX ring_size, const void *opaque, size_t size)                  \
+{                                                                             \
+    if (*masked_prod < masked_cons ||                                         \
+        size <= ring_size - *masked_prod) {                                   \
+        memcpy(buf + *masked_prod, opaque, size);                             \
+    } else {                                                                  \
+        memcpy(buf + *masked_prod, opaque, ring_size - *masked_prod);         \
+        memcpy(buf, (unsigned char *)opaque + (ring_size - *masked_prod),     \
+                size - (ring_size - *masked_prod));                           \
+    }                                                                         \
+    *masked_prod = name##_mask(*masked_prod + size, ring_size);               \
+}                                                                             \
+                                                                              \
+struct name##_data {                                                          \
+    unsigned char *in; /* half of the allocation */                           \
+    unsigned char *out; /* half of the allocation */                          \
+};                                                                            \
+                                                                              \
+                                                                              \
+static inline RING_IDX name##_queued(RING_IDX prod,                           \
+        RING_IDX cons, RING_IDX ring_size)                                    \
+{                                                                             \
+    RING_IDX size;                                                            \
+                                                                              \
+    if (prod == cons)                                                         \
+        return 0;                                                             \
+                                                                              \
+    prod = name##_mask(prod, ring_size);                                      \
+    cons = name##_mask(cons, ring_size);                                      \
+                                                                              \
+    if (prod == cons)                                                         \
+        return ring_size;                                                     \
+                                                                              \
+    if (prod > cons)                                                          \
+        size = prod - cons;                                                   \
+    else                                                                      \
+        size = ring_size - (cons - prod);                                     \
+    return size;                                                              \
+};
+
+#define DEFINE_XEN_FLEX_RING_AND_INTF(name)                                   \
+struct name##_data_intf {                                                     \
+    RING_IDX in_cons, in_prod;                                                \
+                                                                              \
+    uint8_t pad1[56];                                                         \
+                                                                              \
+    RING_IDX out_cons, out_prod;                                              \
+                                                                              \
+    uint8_t pad2[56];                                                         \
+                                                                              \
+    RING_IDX ring_order;                                                      \
+    grant_ref_t ref[];                                                        \
+};                                                                            \
+DEFINE_XEN_FLEX_RING(name);
+
 #endif /* __XEN_PUBLIC_IO_RING_H__ */
 
 /*
-- 
1.9.1


_______________________________________________
Xen-devel mailing list
Xen-devel@lists.xen.org
https://lists.xen.org/xen-devel

[PATCH v5 2/4] xen: introduce a C99 headers check

[Stefano Stabellini]

Introduce a C99 headers check, for non-ANSI compliant headers: 9pfs.h
and pvcalls.h.

In addition to the usual -include stdint.h, also add -include string.h
to the C99 check to get the declaration of memcpy and size_t.

For the same reason, also add -include cstring to the C++ check when
necessary.

Signed-off-by: Stefano Stabellini <stefano@aporeto.com>
CC: JBeulich@suse.com
CC: konrad.wilk@oracle.com
---
 .gitignore           |  1 +
 xen/include/Makefile | 29 +++++++++++++++++++++++++----
 2 files changed, 26 insertions(+), 4 deletions(-)

diff --git a/.gitignore b/.gitignore
index 443b12a..a8905b1 100644
--- a/.gitignore
+++ b/.gitignore
@@ -274,6 +274,7 @@ xen/arch/*/efi/compat.c
 xen/arch/*/efi/efi.h
 xen/arch/*/efi/runtime.c
 xen/include/headers.chk
+xen/include/headers99.chk
 xen/include/headers++.chk
 xen/include/asm
 xen/include/asm-*/asm-offsets.h
diff --git a/xen/include/Makefile b/xen/include/Makefile
index aca7f20..f9d18cd 100644
--- a/xen/include/Makefile
+++ b/xen/include/Makefile
@@ -90,11 +90,16 @@ compat/xlat.h: $(addprefix compat/.xlat/,$(xlat-y)) Makefile
 
 ifeq ($(XEN_TARGET_ARCH),$(XEN_COMPILE_ARCH))
 
-all: headers.chk headers++.chk
+all: headers.chk headers99.chk headers++.chk
 
 PUBLIC_HEADERS := $(filter-out public/arch-% public/dom0_ops.h, $(wildcard public/*.h public/*/*.h) $(public-y))
 
-PUBLIC_ANSI_HEADERS := $(filter-out public/%ctl.h public/xsm/% public/%hvm/save.h, $(PUBLIC_HEADERS))
+PUBLIC_C99_HEADERS :=
+PUBLIC_ANSI_HEADERS := $(filter-out public/%ctl.h public/xsm/% public/%hvm/save.h $(PUBLIC_C99_HEADERS), $(PUBLIC_HEADERS))
+
+EXTRA_PREREQ_C99 := -include string.h
+EXTRA_PREREQ_CPP := -include cstring
+HEADERS_HAVE_EXTRA_PREREQ := $(PUBLIC_C99_HEADERS)
 
 headers.chk: $(PUBLIC_ANSI_HEADERS) Makefile
 	for i in $(filter %.h,$^); do \
@@ -104,12 +109,28 @@ headers.chk: $(PUBLIC_ANSI_HEADERS) Makefile
 	done >$@.new
 	mv $@.new $@
 
+headers99.chk: $(PUBLIC_C99_HEADERS) Makefile
+	for i in $(filter %.h,$^); do \
+	    $(CC) -x c -std=c99 -Wall -Werror -include stdint.h \
+	          $(EXTRA_PREREQ_C99) -S -o /dev/null $$i || exit 1; \
+	    echo $$i; \
+	done >$@.new
+	mv $@.new $@
+
 headers++.chk: $(PUBLIC_HEADERS) Makefile
 	if $(CXX) -v >/dev/null 2>&1; then \
 	    for i in $(filter %.h,$^); do \
+	        for j in $(HEADERS_HAVE_EXTRA_PREREQ); do \
+	            extra="" ; \
+	            if test "$$j" = "$$i"; then \
+	                extra="$(EXTRA_PREREQ_CPP)" ; \
+	                break ; \
+	            fi ; \
+	        done ; \
 	        echo '#include "'$$i'"' \
 	        | $(CXX) -x c++ -std=gnu++98 -Wall -Werror -D__XEN_TOOLS__ \
-	          -include stdint.h -include public/xen.h -S -o /dev/null - \
+	          -include stdint.h $$extra -include public/xen.h \
+	          -S -o /dev/null - \
 	        || exit 1; \
 	        echo $$i; \
 	    done ; \
@@ -128,5 +149,5 @@ all: $(BASEDIR)/include/asm-x86/cpuid-autogen.h
 endif
 
 clean::
-	rm -rf compat headers.chk headers++.chk
+	rm -rf compat headers*.chk
 	rm -f $(BASEDIR)/include/asm-x86/cpuid-autogen.h
-- 
1.9.1


_______________________________________________
Xen-devel mailing list
Xen-devel@lists.xen.org
https://lists.xen.org/xen-devel

Re: [PATCH v5 2/4] xen: introduce a C99 headers check

[Jan Beulich]

>>> On 29.03.17 at 00:08, <sstabellini@kernel.org> wrote:
> Introduce a C99 headers check, for non-ANSI compliant headers: 9pfs.h
> and pvcalls.h.
> 
> In addition to the usual -include stdint.h, also add -include string.h
> to the C99 check to get the declaration of memcpy and size_t.

No, as explained before. You shouldn't think of just your new headers,
but others which may later join the party.

> --- a/xen/include/Makefile
> +++ b/xen/include/Makefile
> @@ -90,11 +90,16 @@ compat/xlat.h: $(addprefix compat/.xlat/,$(xlat-y)) Makefile
>  
>  ifeq ($(XEN_TARGET_ARCH),$(XEN_COMPILE_ARCH))
>  
> -all: headers.chk headers++.chk
> +all: headers.chk headers99.chk headers++.chk
>  
>  PUBLIC_HEADERS := $(filter-out public/arch-% public/dom0_ops.h, $(wildcard 
> public/*.h public/*/*.h) $(public-y))
>  
> -PUBLIC_ANSI_HEADERS := $(filter-out public/%ctl.h public/xsm/% public/%hvm/save.h, $(PUBLIC_HEADERS))
> +PUBLIC_C99_HEADERS :=
> +PUBLIC_ANSI_HEADERS := $(filter-out public/%ctl.h public/xsm/% public/%hvm/save.h $(PUBLIC_C99_HEADERS), $(PUBLIC_HEADERS))
> +
> +EXTRA_PREREQ_C99 := -include string.h
> +EXTRA_PREREQ_CPP := -include cstring

These should be per header, e.g.

pvcalls.h-c99-prereq := string.h
pvcalls.h-cxx-prereq := cstring

which will also (I think at least) greatly simplify the adjustments
needed further down).

Implied from this - please don't use CPP for C++, as CPP is commonly
used for the preprocessor.

Jan


_______________________________________________
Xen-devel mailing list
Xen-devel@lists.xen.org
https://lists.xen.org/xen-devel

Re: [PATCH v5 2/4] xen: introduce a C99 headers check

[Stefano Stabellini]

On Wed, 29 Mar 2017, Jan Beulich wrote:
> >>> On 29.03.17 at 00:08, <sstabellini@kernel.org> wrote:
> > Introduce a C99 headers check, for non-ANSI compliant headers: 9pfs.h
> > and pvcalls.h.
> > 
> > In addition to the usual -include stdint.h, also add -include string.h
> > to the C99 check to get the declaration of memcpy and size_t.
> 
> No, as explained before. You shouldn't think of just your new headers,
> but others which may later join the party.
> 
> > --- a/xen/include/Makefile
> > +++ b/xen/include/Makefile
> > @@ -90,11 +90,16 @@ compat/xlat.h: $(addprefix compat/.xlat/,$(xlat-y)) Makefile
> >  
> >  ifeq ($(XEN_TARGET_ARCH),$(XEN_COMPILE_ARCH))
> >  
> > -all: headers.chk headers++.chk
> > +all: headers.chk headers99.chk headers++.chk
> >  
> >  PUBLIC_HEADERS := $(filter-out public/arch-% public/dom0_ops.h, $(wildcard 
> > public/*.h public/*/*.h) $(public-y))
> >  
> > -PUBLIC_ANSI_HEADERS := $(filter-out public/%ctl.h public/xsm/% public/%hvm/save.h, $(PUBLIC_HEADERS))
> > +PUBLIC_C99_HEADERS :=
> > +PUBLIC_ANSI_HEADERS := $(filter-out public/%ctl.h public/xsm/% public/%hvm/save.h $(PUBLIC_C99_HEADERS), $(PUBLIC_HEADERS))
> > +
> > +EXTRA_PREREQ_C99 := -include string.h
> > +EXTRA_PREREQ_CPP := -include cstring
> 
> These should be per header, e.g.
> 
> pvcalls.h-c99-prereq := string.h
> pvcalls.h-cxx-prereq := cstring
> 
> which will also (I think at least) greatly simplify the adjustments
> needed further down).
> 
> Implied from this - please don't use CPP for C++, as CPP is commonly
> used for the preprocessor.

I can try, but it requires changing the loops below from bash to
Makefile ($(foreach)) to handle variable expansions properly. It's not a
trivial change and it decreases readability, see the new version of the
series that I am about to send. If it was up to me I would probably keep
the code as is in this version.

_______________________________________________
Xen-devel mailing list
Xen-devel@lists.xen.org
https://lists.xen.org/xen-devel

[PATCH v5 3/4] Introduce the Xen 9pfs transport header

[Stefano Stabellini]

Define the ring according to the protocol specification, using the new
DEFINE_XEN_FLEX_RING_AND_INTF macro.

Add the header to the C99 check.

Signed-off-by: Stefano Stabellini <stefano@aporeto.com>
CC: JBeulich@suse.com
CC: konrad.wilk@oracle.com
---
 xen/include/Makefile         |  2 +-
 xen/include/public/io/9pfs.h | 49 ++++++++++++++++++++++++++++++++++++++++++++
 2 files changed, 50 insertions(+), 1 deletion(-)
 create mode 100644 xen/include/public/io/9pfs.h

diff --git a/xen/include/Makefile b/xen/include/Makefile
index f9d18cd..6302ab2 100644
--- a/xen/include/Makefile
+++ b/xen/include/Makefile
@@ -94,7 +94,7 @@ all: headers.chk headers99.chk headers++.chk
 
 PUBLIC_HEADERS := $(filter-out public/arch-% public/dom0_ops.h, $(wildcard public/*.h public/*/*.h) $(public-y))
 
-PUBLIC_C99_HEADERS :=
+PUBLIC_C99_HEADERS := public/io/9pfs.h
 PUBLIC_ANSI_HEADERS := $(filter-out public/%ctl.h public/xsm/% public/%hvm/save.h $(PUBLIC_C99_HEADERS), $(PUBLIC_HEADERS))
 
 EXTRA_PREREQ_C99 := -include string.h
diff --git a/xen/include/public/io/9pfs.h b/xen/include/public/io/9pfs.h
new file mode 100644
index 0000000..4bfd5d4
--- /dev/null
+++ b/xen/include/public/io/9pfs.h
@@ -0,0 +1,49 @@
+/*
+ * 9pfs.h -- Xen 9PFS transport
+ *
+ * Refer to docs/misc/9pfs.markdown for the specification
+ *
+ * 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) 2017 Stefano Stabellini <stefano@aporeto.com>
+ */
+
+#ifndef __XEN_PUBLIC_IO_9PFS_H__
+#define __XEN_PUBLIC_IO_9PFS_H__
+
+#include "../grant_table.h"
+#include "ring.h"
+
+/*
+ * See docs/misc/9pfs.markdown in xen.git for the full specification:
+ * https://xenbits.xen.org/docs/unstable/misc/9pfs.html
+ */
+DEFINE_XEN_FLEX_RING_AND_INTF(xen_9pfs);
+
+#endif
+
+/*
+ * Local variables:
+ * mode: C
+ * c-file-style: "BSD"
+ * c-basic-offset: 4
+ * tab-width: 4
+ * indent-tabs-mode: nil
+ * End:
+ */
-- 
1.9.1


_______________________________________________
Xen-devel mailing list
Xen-devel@lists.xen.org
https://lists.xen.org/xen-devel

[PATCH v5 4/4] Introduce the pvcalls header

[Stefano Stabellini]

Define the ring and request and response structs according to the
specification. Use the new DEFINE_XEN_FLEX_RING macro.

Add the header to the C99 check.

Signed-off-by: Stefano Stabellini <stefano@aporeto.com>
CC: JBeulich@suse.com
CC: konrad.wilk@oracle.com
---
 xen/include/Makefile            |   2 +-
 xen/include/public/io/pvcalls.h | 153 ++++++++++++++++++++++++++++++++++++++++
 2 files changed, 154 insertions(+), 1 deletion(-)
 create mode 100644 xen/include/public/io/pvcalls.h

diff --git a/xen/include/Makefile b/xen/include/Makefile
index 6302ab2..2a37ace 100644
--- a/xen/include/Makefile
+++ b/xen/include/Makefile
@@ -94,7 +94,7 @@ all: headers.chk headers99.chk headers++.chk
 
 PUBLIC_HEADERS := $(filter-out public/arch-% public/dom0_ops.h, $(wildcard public/*.h public/*/*.h) $(public-y))
 
-PUBLIC_C99_HEADERS := public/io/9pfs.h
+PUBLIC_C99_HEADERS := public/io/9pfs.h public/io/pvcalls.h
 PUBLIC_ANSI_HEADERS := $(filter-out public/%ctl.h public/xsm/% public/%hvm/save.h $(PUBLIC_C99_HEADERS), $(PUBLIC_HEADERS))
 
 EXTRA_PREREQ_C99 := -include string.h
diff --git a/xen/include/public/io/pvcalls.h b/xen/include/public/io/pvcalls.h
new file mode 100644
index 0000000..cb81712
--- /dev/null
+++ b/xen/include/public/io/pvcalls.h
@@ -0,0 +1,153 @@
+/*
+ * pvcalls.h -- Xen PV Calls Protocol
+ *
+ * Refer to docs/misc/pvcalls.markdown for the specification
+ *
+ * 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) 2017 Stefano Stabellini <stefano@aporeto.com>
+ */
+
+#ifndef __XEN_PUBLIC_IO_PVCALLS_H__
+#define __XEN_PUBLIC_IO_PVCALLS_H__
+
+#include "../grant_table.h"
+#include "ring.h"
+
+/*
+ * See docs/misc/pvcalls.markdown in xen.git for the full specification:
+ * https://xenbits.xen.org/docs/unstable/misc/pvcalls.html
+ */
+struct pvcalls_data_intf {
+    RING_IDX in_cons, in_prod, in_error;
+
+    uint8_t pad1[52];
+
+    RING_IDX out_cons, out_prod, out_error;
+
+    uint8_t pad2[52];
+
+    RING_IDX ring_order;
+    grant_ref_t ref[];
+};
+DEFINE_XEN_FLEX_RING(pvcalls);
+
+#define PVCALLS_SOCKET         0
+#define PVCALLS_CONNECT        1
+#define PVCALLS_RELEASE        2
+#define PVCALLS_BIND           3
+#define PVCALLS_LISTEN         4
+#define PVCALLS_ACCEPT         5
+#define PVCALLS_POLL           6
+
+struct xen_pvcalls_request {
+    uint32_t req_id; /* private to guest, echoed in response */
+    uint32_t cmd;    /* command to execute */
+    union {
+        struct xen_pvcalls_socket {
+            uint64_t id;
+            uint32_t domain;
+            uint32_t type;
+            uint32_t protocol;
+        } socket;
+        struct xen_pvcalls_connect {
+            uint64_t id;
+            uint8_t addr[28];
+            uint32_t len;
+            uint32_t flags;
+            grant_ref_t ref;
+            uint32_t evtchn;
+        } connect;
+        struct xen_pvcalls_release {
+            uint64_t id;
+            uint8_t reuse;
+        } release;
+        struct xen_pvcalls_bind {
+            uint64_t id;
+            uint8_t addr[28];
+            uint32_t len;
+        } bind;
+        struct xen_pvcalls_listen {
+            uint64_t id;
+            uint32_t backlog;
+        } listen;
+        struct xen_pvcalls_accept {
+            uint64_t id;
+            uint64_t id_new;
+            grant_ref_t ref;
+            uint32_t evtchn;
+        } accept;
+        struct xen_pvcalls_poll {
+            uint64_t id;
+        } poll;
+        /* dummy member to force sizeof(struct xen_pvcalls_request)
+         * to match across archs */
+        struct xen_pvcalls_dummy {
+            uint8_t dummy[56];
+        } dummy;
+    } u;
+};
+
+struct xen_pvcalls_response {
+    uint32_t req_id;
+    uint32_t cmd;
+    int32_t ret;
+    uint32_t pad;
+    union {
+        struct _xen_pvcalls_socket {
+            uint64_t id;
+        } socket;
+        struct _xen_pvcalls_connect {
+            uint64_t id;
+        } connect;
+        struct _xen_pvcalls_release {
+            uint64_t id;
+        } release;
+        struct _xen_pvcalls_bind {
+            uint64_t id;
+        } bind;
+        struct _xen_pvcalls_listen {
+            uint64_t id;
+        } listen;
+        struct _xen_pvcalls_accept {
+            uint64_t id;
+        } accept;
+        struct _xen_pvcalls_poll {
+            uint64_t id;
+        } poll;
+        struct _xen_pvcalls_dummy {
+            uint8_t dummy[8];
+        } dummy;
+    } u;
+};
+
+DEFINE_RING_TYPES(xen_pvcalls, struct xen_pvcalls_request,
+                  struct xen_pvcalls_response);
+
+#endif
+
+/*
+ * Local variables:
+ * mode: C
+ * c-file-style: "BSD"
+ * c-basic-offset: 4
+ * tab-width: 4
+ * indent-tabs-mode: nil
+ * End:
+ */
-- 
1.9.1


_______________________________________________
Xen-devel mailing list
Xen-devel@lists.xen.org
https://lists.xen.org/xen-devel

Re: [PATCH v5 1/4] ring.h: introduce macros to handle monodirectional rings with multiple req sizes

[Jan Beulich]

>>> On 29.03.17 at 00:08, <sstabellini@kernel.org> wrote:
> +#define DEFINE_XEN_FLEX_RING(name)                                            \
> +static inline RING_IDX name##_mask(RING_IDX idx, RING_IDX ring_size)          \
> +{                                                                             \
> +    return (idx & (ring_size - 1));                                           \
> +}                                                                             \
> +                                                                              \
> +static inline RING_IDX name##_mask_order(RING_IDX idx, RING_IDX ring_order)   \
> +{                                                                             \
> +    return (idx & (XEN_FLEX_RING_SIZE(ring_order) - 1));                      \
> +}                                                                             \

Do you really need both (and if you do, perhaps the latter should
call the former)? I also find the mixture of ring_order and ring_size
parameters of later functions a little strange.

> +static inline unsigned char *name##_get_ring_ptr(unsigned char *buf,          \
> +                                                 RING_IDX idx,                \
> +                                                 RING_IDX ring_order)         \
> +{                                                                             \
> +    return buf + name##_mask_order(idx, ring_order);                          \

Please be consistent with parenthesizing the operand of return:
The earlier two functions have an unnecessary pair of parens,
so personally I'd prefer those to be dropped. But if you prefer to
have them, add them everywhere.

> +static inline void name##_read_packet(const unsigned char *buf,               \
> +        RING_IDX masked_prod, RING_IDX *masked_cons,                          \
> +        RING_IDX ring_size, void *opaque, size_t size)                        \

Especially with so many parameters I think some extra thought
should be spent on their ordering: Primarily this is a memcpy()-
like function, so I would kind of expect destination description,
source description (each of which may require more than one
parameter), size, auxiliary. As to the auxiliary part (especially
ring_size) - there's no structure you could pass a pointer to,
taking care of more than one of these, is there (struct
name##_data_intf would at least appear to be a candidate,
but is not always available)?

I'm also not really clear whether it wouldn't be better for both
input and output to be void * (input remaining const of course).

And finally please indent function parameter declarations
uniformly throughout the patch. If you prefer to follow the style
of the declaration right above, then please reduce indentation 
to four spaces (to match that of function scope local variable
declarations).

> +static inline RING_IDX name##_queued(RING_IDX prod,                           \
> +        RING_IDX cons, RING_IDX ring_size)                                    \
> +{                                                                             \
> +    RING_IDX size;                                                            \
> +                                                                              \
> +    if (prod == cons)                                                         \
> +        return 0;                                                             \
> +                                                                              \
> +    prod = name##_mask(prod, ring_size);                                      \
> +    cons = name##_mask(cons, ring_size);                                      \
> +                                                                              \
> +    if (prod == cons)                                                         \
> +        return ring_size;                                                     \
> +                                                                              \
> +    if (prod > cons)                                                          \
> +        size = prod - cons;                                                   \
> +    else                                                                      \
> +        size = ring_size - (cons - prod);                                     \
> +    return size;                                                              \
> +};

Stray semicolon at end of function definition.

> +#define DEFINE_XEN_FLEX_RING_AND_INTF(name)                                   \
> +struct name##_data_intf {                                                     \
> +    RING_IDX in_cons, in_prod;                                                \
> +                                                                              \
> +    uint8_t pad1[56];                                                         \
> +                                                                              \
> +    RING_IDX out_cons, out_prod;                                              \
> +                                                                              \
> +    uint8_t pad2[56];                                                         \
> +                                                                              \
> +    RING_IDX ring_order;                                                      \
> +    grant_ref_t ref[];                                                        \
> +};                                                                            \
> +DEFINE_XEN_FLEX_RING(name);

The trailing semicolon here should be left out, requiring the use site
to put one after the macro invocation. Some compilers warn about
such stray semicolons. This implies that the last element of 
DEFINE_XEN_FLEX_RING() should also not be an inline function
definition (as a semicolon placed after the macro invocation would
then also possibly trigger a compiler warning).

Jan

_______________________________________________
Xen-devel mailing list
Xen-devel@lists.xen.org
https://lists.xen.org/xen-devel

Re: [PATCH v5 1/4] ring.h: introduce macros to handle monodirectional rings with multiple req sizes

[Stefano Stabellini]

On Wed, 29 Mar 2017, Jan Beulich wrote:
> >>> On 29.03.17 at 00:08, <sstabellini@kernel.org> wrote:
> > +#define DEFINE_XEN_FLEX_RING(name)                                            \
> > +static inline RING_IDX name##_mask(RING_IDX idx, RING_IDX ring_size)          \
> > +{                                                                             \
> > +    return (idx & (ring_size - 1));                                           \
> > +}                                                                             \
> > +                                                                              \
> > +static inline RING_IDX name##_mask_order(RING_IDX idx, RING_IDX ring_order)   \
> > +{                                                                             \
> > +    return (idx & (XEN_FLEX_RING_SIZE(ring_order) - 1));                      \
> > +}                                                                             \
> 
> Do you really need both (and if you do, perhaps the latter should
> call the former)? I also find the mixture of ring_order and ring_size
> parameters of later functions a little strange.

Actually, I don't need two, I'll drop name##_mask_order. I'll change
the parameter below to be ring_size for consistency.


> > +static inline unsigned char *name##_get_ring_ptr(unsigned char *buf,          \
> > +                                                 RING_IDX idx,                \
> > +                                                 RING_IDX ring_order)         \
> > +{                                                                             \
> > +    return buf + name##_mask_order(idx, ring_order);                          \
> 
> Please be consistent with parenthesizing the operand of return:
> The earlier two functions have an unnecessary pair of parens,
> so personally I'd prefer those to be dropped. But if you prefer to
> have them, add them everywhere.

OK


> > +static inline void name##_read_packet(const unsigned char *buf,               \
> > +        RING_IDX masked_prod, RING_IDX *masked_cons,                          \
> > +        RING_IDX ring_size, void *opaque, size_t size)                        \
> 
> Especially with so many parameters I think some extra thought
> should be spent on their ordering: Primarily this is a memcpy()-
> like function, so I would kind of expect destination description,
> source description (each of which may require more than one
> parameter), size, auxiliary.

I'll reorder the parameters.


> As to the auxiliary part (especially
> ring_size) - there's no structure you could pass a pointer to,
> taking care of more than one of these, is there (struct
> name##_data_intf would at least appear to be a candidate,
> but is not always available)?

That is the problem, it is not always available. I prefer to keep them
separate.


> I'm also not really clear whether it wouldn't be better for both
> input and output to be void * (input remaining const of course).

I think it's a matter of taste: source is unsigned char* because it is
of the same type as the underlying ring buffer to read data from. I'll
leave it as is for now.


> And finally please indent function parameter declarations
> uniformly throughout the patch. If you prefer to follow the style
> of the declaration right above, then please reduce indentation 
> to four spaces (to match that of function scope local variable
> declarations).

I'll fix indentation.


> > +static inline RING_IDX name##_queued(RING_IDX prod,                           \
> > +        RING_IDX cons, RING_IDX ring_size)                                    \
> > +{                                                                             \
> > +    RING_IDX size;                                                            \
> > +                                                                              \
> > +    if (prod == cons)                                                         \
> > +        return 0;                                                             \
> > +                                                                              \
> > +    prod = name##_mask(prod, ring_size);                                      \
> > +    cons = name##_mask(cons, ring_size);                                      \
> > +                                                                              \
> > +    if (prod == cons)                                                         \
> > +        return ring_size;                                                     \
> > +                                                                              \
> > +    if (prod > cons)                                                          \
> > +        size = prod - cons;                                                   \
> > +    else                                                                      \
> > +        size = ring_size - (cons - prod);                                     \
> > +    return size;                                                              \
> > +};
> 
> Stray semicolon at end of function definition.

OK


> > +#define DEFINE_XEN_FLEX_RING_AND_INTF(name)                                   \
> > +struct name##_data_intf {                                                     \
> > +    RING_IDX in_cons, in_prod;                                                \
> > +                                                                              \
> > +    uint8_t pad1[56];                                                         \
> > +                                                                              \
> > +    RING_IDX out_cons, out_prod;                                              \
> > +                                                                              \
> > +    uint8_t pad2[56];                                                         \
> > +                                                                              \
> > +    RING_IDX ring_order;                                                      \
> > +    grant_ref_t ref[];                                                        \
> > +};                                                                            \
> > +DEFINE_XEN_FLEX_RING(name);
> 
> The trailing semicolon here should be left out, requiring the use site
> to put one after the macro invocation. Some compilers warn about
> such stray semicolons. This implies that the last element of 
> DEFINE_XEN_FLEX_RING() should also not be an inline function
> definition (as a semicolon placed after the macro invocation would
> then also possibly trigger a compiler warning).

OK

_______________________________________________
Xen-devel mailing list
Xen-devel@lists.xen.org
https://lists.xen.org/xen-devel

Re: [PATCH v5 1/4] ring.h: introduce macros to handle monodirectional rings with multiple req sizes

[Jan Beulich]

>>> On 29.03.17 at 00:08, <sstabellini@kernel.org> wrote:
> --- a/xen/include/public/io/ring.h
> +++ b/xen/include/public/io/ring.h
> @@ -27,6 +27,21 @@
>  #ifndef __XEN_PUBLIC_IO_RING_H__
>  #define __XEN_PUBLIC_IO_RING_H__
>  
> +/*
> + * When #include'ing this header, you need to provide the following
> + * declaration upfront:
> + * - standard integers types (uint8_t, uint16_t, etc)
> + * They are provided by stdint.h of the standard headers.
> + *
> + * In addition, if you intend to use the FLEX macros, you also need to
> + * provide:
> + * - size_t
> + * - memcpy
> + * - grant_ref_t
> + * These declarations are provided by string.h of the standard headers,
> + * and grant_table.h from the Xen public headers.
> + */

Btw, there's another difference only implied so far: The integer
types indeed need to be made available up front (as spelled out).
The others can as well be made available after ring.h inclusion,
as long as that happens before invoking any of the macros.

Jan


_______________________________________________
Xen-devel mailing list
Xen-devel@lists.xen.org
https://lists.xen.org/xen-devel

Re: [PATCH v5 1/4] ring.h: introduce macros to handle monodirectional rings with multiple req sizes

[Stefano Stabellini]

On Wed, 29 Mar 2017, Jan Beulich wrote:
> >>> On 29.03.17 at 00:08, <sstabellini@kernel.org> wrote:
> > --- a/xen/include/public/io/ring.h
> > +++ b/xen/include/public/io/ring.h
> > @@ -27,6 +27,21 @@
> >  #ifndef __XEN_PUBLIC_IO_RING_H__
> >  #define __XEN_PUBLIC_IO_RING_H__
> >  
> > +/*
> > + * When #include'ing this header, you need to provide the following
> > + * declaration upfront:
> > + * - standard integers types (uint8_t, uint16_t, etc)
> > + * They are provided by stdint.h of the standard headers.
> > + *
> > + * In addition, if you intend to use the FLEX macros, you also need to
> > + * provide:
> > + * - size_t
> > + * - memcpy
> > + * - grant_ref_t
> > + * These declarations are provided by string.h of the standard headers,
> > + * and grant_table.h from the Xen public headers.
> > + */
> 
> Btw, there's another difference only implied so far: The integer
> types indeed need to be made available up front (as spelled out).
> The others can as well be made available after ring.h inclusion,
> as long as that happens before invoking any of the macros.

OK

_______________________________________________
Xen-devel mailing list
Xen-devel@lists.xen.org
https://lists.xen.org/xen-devel