Currently the layout of a struct is used to compute its size. This works
fine in case of structs of fixed size. However this introduces
forwards-compatibility problems in cases when the struct has multiple
variants and the exact variant is specified by the value of some field
(e.g. in the case of <switch> elements). Future revisions of protocols
may introduce new layout variants, in which case the old code does not
know the size of the struct variant and can't parse the incoming byte
stream.

Instead of relying on knowledge about the layout of data structures we
should instead use the length field for length information. This way
when old client libxcb communicates with newer server it can at least
ignore unknown struct variants.

The Xserver itself is capable of terminating itself once all X11 clients
are gone, yet in a typical full session, there are a number of X11
clients running continuously (e.g. the Xsettings daemon, IBus, etc.).

Those always-running clients will prevent the Xserver from terminating,
because the actual number of X11 clients will never drop to 0.

To solve this issue directly at the Xserver level, this add new entries
to the XFixes extension to let the X11 clients themselves specify the
disconnect mode they expect.

Typically, those X11 daemon clients would specify the disconnect mode
XFixesClientDisconnectFlagTerminate to let the Xserver know that they
should not be accounted for when checking the remaining clients prior
to terminate.

Signed-off-by: Olivier Fourdan <ofourdan@redhat.com>
Reviewed-by: Simon Ser <contact@emersion.fr>

A little too much copy-and-paste from the GrabPointer doc text had
happened here.

Signed-off-by: Alan Coopersmith <alan.coopersmith@oracle.com>

Signed-off-by: Alan Coopersmith <alan.coopersmith@oracle.com>

Signed-off-by: Samanta Navarro <ferivoz@riseup.net>

In commit 7d58eed9 we started
to use xml.etree.ElementTree for Python >= 3.9.  In fact the
xml.etree.cElementTree module has already been deprecated
since Python 3.3.

Given this fact, we should start to use the xml.etree.ElementTree
module beginning with Python 3.3.

See: https://github.com/python/cpython/commit/a72a98f24a19928e31dcc4cab2cd2ad0f1846e11



Signed-off-by: Björn Esser <besser82@fedoraproject.org>

It can be replaced with xml.etree.ElementTree safely.

Signed-off-by: Björn Esser <besser82@fedoraproject.org>

fractions.gcd() has been deprecated since Python 3.5, and
was finally dropped in Python 3.9.  It is recommended to
use math.gcd() instead.

Signed-off-by: Björn Esser <besser82@fedoraproject.org>

In xv.xml, there is something like this:

  <struct name="ImageFormatInfo">
    [...]
    <field type="CARD8" name="byte_order" enum="ImageOrder" />
    <pad bytes="2" />
    <list type="CARD8" name="guid">
      <value>16</value>
    </list>
    [...]
  </struct>

When parsing this, the Field instance for "guid" ended up with .enum ==
"ImageOrder". This is because the loop that parses complex type did not
unset a variable across iterations, meaning that the last "enum"
property "stuck" and was also used for all following fields.

Fix this by simply moving the initialisation of the "enum" variable
inside of the loop.

Signed-off-by: Uli Schlachter <psychon@znc.in>

Signed-off-by: Matt Turner <mattst88@gmail.com>

Signed-off-by: Matt Turner <mattst88@gmail.com>

xcbgen 'helpfully' transforms things to C types already so that libxcb
does not have to do so. Thus, even though the XML says that a field has
type CARD8, xcbgen will claim uint8_t. This might be a bit weird, but is
so far totally fine.

However, the type mapping that xcbgen uses is not injective. All of
CARD8, BYTE and BOOL get turned into uint8_t and the original type is
lost.

This is totally fine for libxcb, but programming languages other than C
do have built in boolean types. My personal problem is with Rust, where
providing a boolean for an integer argument causes a compiler error.
This results in (relatively) ugly "0 / 1" instead of "false / true".

This commit adds a new xml_type member to SimpleType. This type contains
the original string that appeared in the XML file.

Since libxcb creates instances of SimpleType itself and to avoid
breaking the API, the new argument to SimpleType.__init__ defaults to
None.

Signed-off-by: Uli Schlachter <psychon@znc.in>

XFixes contains a CreateRegion request:

   <request name="CreateRegion" opcode="5">
     <field type="REGION"    name="region" />
     <list  type="RECTANGLE" name="rectangles" />
   </request>

This request contains a list of type RECTANGLE. This struct comes from
xproto and is thus not contained in xfixes itself.

Normal "Struct"s have their resolve() method called early, because they
appear in the module itself. However, in the CreateRegion case, this
struct is imported and thus does not get resolved. Instead, ListType's
resolve() method calls self.member.resolve(module). Thus, only at this
point is the struct resolved.

Why is this important? Struct.resolve() is the same as
ComplexType.resolve() and this function does self.calc_size() at the
end. Thus, only after the struct was resolved is its size known. Before
that, the size is just set to 0 (this happens in ComplexType.__init__).

However, ListType.__init__ already computes its size member based on its
member. At this point, this is still 0 so the list ends up believing its
size to be zero.

Fix this by recomputing self.size in ListType.resolve().

Signed-off-by: Uli Schlachter <psychon@znc.in>

According to git grep '\.fds', this does not appear anywhere else in
xcb-proto or libxcb.

Signed-off-by: Uli Schlachter <psychon@znc.in>

Daniel Stone authored 7 years ago and Jon Turney committed 5 years ago

xcb-proto's pkg-config file exports static file paths for where to find
its Python files and the XML definitions.

When used to discover cflags, library paths, etc, pkg-config will
prepend ${PKG_CONFIG_SYSROOT_DIR}, if any, to those paths. This makes
the use of a sysroot prefix transparent to users. However, since it
doesn't know about paths in custom variables, it cannot automatically
prefix this.

Adding ${pc_sysrootdir} to these absolute paths makes the behaviour
match, and XCB builds work out of the box in a sysroot.

Signed-off-by: Daniel Stone <daniels@collabora.com>
Reviewed-by: Eric Anholt <eric@anholt.net>

Signed-off-by: Alan Coopersmith <alan.coopersmith@oracle.com>

This was set to BOOL, but the protocol headers used Bool, presumably a
32-bit type. We're switching everything to CARD32 as the best option
for compatibility.

Signed-off-by: Keith Packard <keithp@keithp.com>
Reviewed-by: Mihai Moldovan <ionic@ionic.de>

This release adds support for variable-sized lists of FDs (e.g. sized
by a field in the request/event), as well as several other cleanups and
fixes in the core code.

DRI3 has been bumped to v1.2, using the new support to support
multi-planar buffers and DRM modifiers.

Present has also been bumped to v1.2, adding a new token for the server
to indicate to the client that it should reallocate its buffers.

RandR has been updated to 1.6, adding support for DRM leases to allow
clients to directly control outputs.

Signed-off-by: Daniel Stone <daniels@collabora.com>

Louis-Francis Ratté-Boulianne authored 7 years ago and Daniel Stone committed 6 years ago

This flag (returned with PresentCompleteNotify event) indicates
that the window could have been flipped instead of copied if
the buffer format/modifier was different. Clients must indicate
that they support the flag by adding option PresentOptionSuboptimal
when calling PresentPixmap.

v2: Add PresentOptionSuboptimal

Signed-off-by: Louis-Francis Ratté-Boulianne <lfrb@collabora.com>

Louis-Francis Ratté-Boulianne authored 7 years ago and Daniel Stone committed 6 years ago

Bumping to version 1.2, add support for:
  - querying formats and modifiers supported by the server
  - creating (and receiving) multi-planar buffers
  - creating (and receiving) buffers with modifiers

v2: Fix alignment
v3: Bump to version 1.2 rather than 1.1. The old dri3proto had a '1.1'
    release for non-functional changes, so we avoid that version and
    skip directly to 1.2.

Signed-off-by: Daniel Stone <daniels@collabora.com>
Signed-off-by: Louis-Francis Ratté-Boulianne <lfrb@collabora.com>

See the description in the randrproto repository.

Add lease notify event

Signed-off-by: Keith Packard <keithp@keithp.com>

Add a special case in ListType to support lists of file descriptors,
which also requires explicit support within clients.

File descriptors are not supported as children of other complex types,
e.g. switch.

Signed-off-by: Daniel Stone <daniels@collabora.com>

The last item from this TODO list has been implemented
by the commit "SendExtensionEvent uses eventstruct":
https://cgit.freedesktop.org/xcb/proto/commit/?id=97fbba25baf241cbcda882739db423b1d36e440a



Signed-off-by: Christian Linhart <chris@demorecorder.com>

Use the eventstruct element to define the SendExtensionEvent request
instead of using a byte-array.

This has the advantage that the structure of the data to be sent
is now defined in the xml definition.
This can be used by generators to make sending events safer.
Such as avoiding type-casts in C.

It can also be useful for protocol analyzers and tools like that
because the structure of the sent events is now defined in the xml-definition.

Tested-by: Christian Linhart <chris@demorecorder.com>
Signed-off-by: Christian Linhart <chris@demorecorder.com>

eventstruct is a new xml element that allows to use events
as part of requests.

This is, e.g., needed by the SendExtensionEvent request
of the XINPUT extension.

Signed-off-by: Christian Linhart <chris@demorecorder.com>

member event_type_base in struct InputClassInfo is a CARD8.
This patch replaces the type with the new typedef EventTypeBase
which is defined as a CARD8.

This does not alter the ABI or API for that struct because
typedefs do not appear in the ABI.
And C handles typedefs as aliases.
So there'll be no API issues as well.

The libxcb generator adds additional functions that implement
an iterator over this type.
This is harmless because adding code is ABI/API compatible.

The purpose of this change is to add more semantic info to this
member variable of this struct.

This helps with xcb-based generators or automatic code-checkers
and things like that.

Signed-off-by: Christian Linhart <chris@demorecorder.com>

Set the lenfield of a sumof expression object, so that
libxcb or other generators won't need to look it up.

This object is trivially available in the parser but
needs a complex lookup in generators.

Signed-off-by: Christian Linhart <chris@demorecorder.com>

aarzilli authored 9 years ago and Christian Linhart committed 7 years ago

Fields  AuthorizationProtocolName and AuthorizationProtocolData of
SetupRequest should be padded:

http://www.x.org/releases/current/doc/xproto/x11protocol.html#Encoding::Connection_Setup

The problem was discovered by github user pphaneuf while trying to use xgb
to write his own implementation of the connection handshake. Neither xgb
nor xcb actually use code generated for SetupRequest for the handshake,
which is why this bug went unnoticed.

https://github.com/BurntSushi/xgb/issues/24



Alessandro Arzilli.

Reviewed-by: Christian Linhart <chris@demorecorder.com>

Jon Turney authored 8 years ago and Uli Schlachter committed 8 years ago

Here is an attempt at updating the schema to add serialize attribute to pad
element and required_start_align element.

Not sure if I've added required_start_align element in the right place. The
default case in the switch element is removed as it doesn't seem to be used, and
otherwise makes the schema ambiguous.

$ make check
Making check in src
make[1]: Entering directory '/jhbuild/x86_64-pc-cygwin/build/xcb/proto/src'
make  check-local
make[2]: Entering directory '/jhbuild/x86_64-pc-cygwin/build/xcb/proto/src'
/usr/bin/xmllint --noout --schema /jhbuild/checkout/xcb/proto/src/xcb.xsd
/jhbuild/checkout/xcb/proto/src/*.xml
/jhbuild/checkout/xcb/proto/src/bigreq.xml validates
/jhbuild/checkout/xcb/proto/src/composite.xml validates
/jhbuild/checkout/xcb/proto/src/damage.xml validates
/jhbuild/checkout/xcb/proto/src/dpms.xml validates
/jhbuild/checkout/xcb/proto/src/dri2.xml validates
/jhbuild/checkout/xcb/proto/src/dri3.xml validates
/jhbuild/checkout/xcb/proto/src/ge.xml validates
Element 'required_start_align': This element is not expected. Expected is one of ( pad, field, list, fd ).
Element 'required_start_align': This element is not expected. Expected is one of ( pad, field, list, fd ).
Element 'required_start_align': This element is not expected. Expected is one of ( pad, field, list, fd ).
Element 'required_start_align': This element is not expected. Expected is one of ( pad, field, list, fd ).
/jhbuild/checkout/xcb/proto/src/glx.xml fails to validate
Element 'required_start_align': This element is not expected. Expected is one of ( pad, field, list, fd, exprfield, switch, reply, doc ).
Element 'required_start_align': This element is not expected. Expected is one of ( pad, field, list, fd, exprfield, switch, reply, doc ).
Element 'required_start_align': This element is not expected. Expected is one of ( pad, field, list, fd, doc ).
Element 'required_start_align': This element is not expected. Expected is one of ( pad, field, list, fd, doc ).
/jhbuild/checkout/xcb/proto/src/present.xml fails to validate
/jhbuild/checkout/xcb/proto/src/randr.xml validates
/jhbuild/checkout/xcb/proto/src/record.xml validates
/jhbuild/checkout/xcb/proto/src/render.xml validates
/jhbuild/checkout/xcb/proto/src/res.xml validates
/jhbuild/checkout/xcb/proto/src/screensaver.xml validates
/jhbuild/checkout/xcb/proto/src/shape.xml validates
/jhbuild/checkout/xcb/proto/src/shm.xml validates
/jhbuild/checkout/xcb/proto/src/sync.xml validates
/jhbuild/checkout/xcb/proto/src/xc_misc.xml validates
/jhbuild/checkout/xcb/proto/src/xevie.xml validates
/jhbuild/checkout/xcb/proto/src/xf86dri.xml validates
/jhbuild/checkout/xcb/proto/src/xf86vidmode.xml validates
/jhbuild/checkout/xcb/proto/src/xfixes.xml validates
/jhbuild/checkout/xcb/proto/src/xinerama.xml validates
Element 'required_start_align': This element is not expected. Expected is one of ( bitcase, case, pad, field, list, fd ).
Element 'required_start_align': This element is not expected. Expected is one of ( bitcase, case, pad, field, list, fd ).
Element 'required_start_align': This element is not expected. Expected is one of ( bitcase, case, pad, field, list, fd ).
/jhbuild/checkout/xcb/proto/src/xinput.xml fails to validate
Element 'pad', attribute 'serialize': The attribute 'serialize' is not allowed.
Element 'pad', attribute 'serialize': The attribute 'serialize' is not allowed.
Element 'pad', attribute 'serialize': The attribute 'serialize' is not allowed.
Element 'pad', attribute 'serialize': The attribute 'serialize' is not allowed.
Element 'pad', attribute 'serialize': The attribute 'serialize' is not allowed.
/jhbuild/checkout/xcb/proto/src/xkb.xml fails to validate
/jhbuild/checkout/xcb/proto/src/xprint.xml validates
/jhbuild/checkout/xcb/proto/src/xproto.xml validates
/jhbuild/checkout/xcb/proto/src/xselinux.xml validates
/jhbuild/checkout/xcb/proto/src/xtest.xml validates
/jhbuild/checkout/xcb/proto/src/xv.xml validates
/jhbuild/checkout/xcb/proto/src/xvmc.xml validates
Makefile:534: recipe for target 'check-local' failed

Signed-off-by: Jon Turney <jon.turney@dronecode.org.uk>
Bugzilla: https://bugs.freedesktop.org/show_bug.cgi?id=95494


Reviewd-by: Ran Benita <ran234@gmail.com>
Signed-off-by: Uli Schlachter <psychon@znc.in>

The specification disagrees with itself, so use the part of the
specification that matches the other implementations.

Reviewed-by: Ran Benita <ran234@gmail.com>
Signed-off-by: Peter Harris <pharris@opentext.com>

Thomas Klausner authored 8 years ago and Uli Schlachter committed 8 years ago

Fixes build with python-3.x.

Signed-off-by: Thomas Klausner <wiz@NetBSD.org>
Signed-off-by: Uli Schlachter <psychon@znc.in>

Thomas Klausner authored 8 years ago and Uli Schlachter committed 8 years ago

At least python-3.5.x complains about this forcefully.

Signed-off-by: Thomas Klausner <wiz@NetBSD.org>
Signed-off-by: Uli Schlachter <psychon@znc.in>

There were more than 100 commits since the last release. The NEWS file only
provides an overview of these changes.

Signed-off-by: Uli Schlachter <psychon@znc.in>

Signed-off-by: Christian Linhart <chris@demorecorder.com>

From now on, due to a patch in libxcb, pads will not be serialized
anymore. This is to maintain ABI-compatibility when adding
explicit align pads.

However, some align pads were already be serialized in prior
official versions of libxcb. Therefore we need a method to
enforce serialization, so we can maintainn ABI compatibility
with that legacy.

Signed-off-by: Christian Linhart <chris@demorecorder.com>

Jaya Tiwari authored 9 years ago and Christian Linhart committed 9 years ago

Some rework done by Christian Linhart

Signed-off-by: Jaya Tiwari <tiwari.jaya18@gmail.com>
Signed-off-by: Christian Linhart <chris@demorecorder.com>

These are just minimal adjustments to get xkb through
the checks of the alignment checker.

It is not the big fixup which I have already posted an RFC patch
a while ago.

V2 of this patch:
  make indentation consistent with the file
  (tabs vs spaces)

Signed-off-by: Christian Linhart <chris@demorecorder.com>

The verification algorithm basically traverses the protocol-description
of a protocol entity by recursively processing all fields and their
types.

A start-align consists of two numbers:
* the alignment: This is a power of 2, and guarantees that the
  address (or protocol position) minus the offset can be divided
  by this number.

* the offset: how many bytes the current position is after a
  position that can be divided by the alignment.

The algorithm starts with the start-alignment and computes the alignment
of each field from the start-align of the previous field as follows:
* if the previous field is a primitive type then the offset is increased
  by the size of the primitive type module the alignment.
  If the alignment or offset are incompatible with the primitive type,
  then an error is reported.
* if the previous field is a complex type, then it is processed recursively.
* if the previous field is an alignment-pad, then the alignment is
  adjusted accordingly, as to be expected by the alignment-pad.

Start-aligns can also be set manually with the xml-element
"required_start_align" which has two attributes: "align" and "offset"
If "offset" is omitted, it is assumed to 0.

All toplevel protocol entities default to 4-byte start-alignment with offset 0.

All non-toplevel complex entities, such as structs, switch, case, ...
do not have a default alignment.
If no alignment is explicitly specified for them, their alignment
is computed by their usage in other entities.
In that case, if they are used with aligments that violate the
alignment requirements of some of their fields, an error is issued.

If they are used with an alignment with non-zero offset,
a warning is issued, which recommends to specify the required_start_align
explicitly. (Reason: I don't want non-zero offsets to be silently
computed automatically. These non-zero offsets have to be reviewed
by a human and specified explicitely to record that this was reviewed
by a human)

If the required_start_align is explicitly specified for an entity
then an error will be issued if it is used with an aligment that's
incompatible with the explicitly specified alignment.

If an entity is used in different contexts with different start-aligns
then those start-aligns are combined to an align which is compatible
with all aligns.
E.g. (align 4, offset 0) and (align 4, offset 2) are combined
to (align 2, offset 0).

Error reports include the relevant context for a misalignment.
For non-toplevel entities this includes the entity-usage stack.
There is some complexity in the algorithm for reducing the size
of the error-reports to include only the relevant info.

This alignment verifier is also a prerequisite for getting
rid of implicit alignment altogether.
(This will then simplify the generated code and make it faster.)

Signed-off-by: Christian Linhart <chris@demorecorder.com>

Especially, these start-aligns have non-zero offsets.

Signed-off-by: Christian Linhart <chris@demorecorder.com>