Other input drivers already shipping pkgconfig files use this naming scheme
and since the 1.8 release didn't install the file anyway, renaming is still
acceptable at this point.
Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net>
Reviewed-by: Alan Coopersmith <alan.coopersmith@oracle.com>
pnp.c: In function 'probePs2ProtocolPnP':
pnp.c:711:31: warning: declaration of 'seq' shadows a previous local
[-Wshadow]
pnp.c:705:23: warning: shadowed declaration is here [-Wshadow]
mouse.c: In function 'MouseCommonOptions':
mouse.c:330:13: warning: declaration of 'i' shadows a previous local
[-Wshadow]
mouse.c:279:9: warning: shadowed declaration is here [-Wshadow]
mouse.c: In function 'autoProbeMouse':
mouse.c:3602:29: warning: declaration of 'proto' shadows a global
declaration [-Wshadow]
mouse.c:2482:22: warning: shadowed declaration is here [-Wshadow]
Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net>
Reviewed-by: Jeremy Huddleston Sequoia <jeremyhu@apple.com>
Reviewed-by: Alan Coopersmith <alan.coopersmith@oracle.com>
Signed-off-by: Alan Coopersmith <alan.coopersmith@oracle.com>
Not all of the button label settings were wrapped in ABI #ifdefs,
so just drop all #ifdefs for GET_ABI_MAJOR(ABI_XINPUT_VERSION) < 7
Signed-off-by: Alan Coopersmith <alan.coopersmith@oracle.com>
Reviewed-by: Peter Hutterer <peter.hutterer@who-t.net>
Reviewed-by: Jeremy Huddleston Sequoia <jeremyhu@apple.com>
Existing code had a mix of places tab & space characters were used.
Make them all spaces for consistency, since that's the new style
chosen for xorg-server.
"git diff -w" shows no changes - this is pure whitespace adjustment.
Signed-off-by: Alan Coopersmith <alan.coopersmith@oracle.com>
cleanup done via: perl -p -i -e 's/[\t ]*$//'
"git diff -w" shows no changes - this is pure whitespace adjustment.
Signed-off-by: Alan Coopersmith <alan.coopersmith@oracle.com>
Based on evdev's similar properties, including using the name "middle"
button, to avoid confusion with evdev's 3rd button emulation for
emulating the right button on a single button mouse.
Allows manual enable & disable at runtime.
Exports new xf86-mouse.pc & xf86-mouse-properties.h for property name
definitions.
Signed-off-by: Alan Coopersmith <alan.coopersmith@oracle.com>
Reviewed-by: Alexandr Shadchin <alexandr.shadchin@gmail.com>
Reviewed-by: Peter Hutterer <peter.hutterer@who-t.net>
Previously they checked pInfo->options, then fell back to
pInfo->conf_idev->commonOptions - but then 7bf22a368c
replaced the latter with the former, resulting in some redunancy.
Signed-off-by: Alan Coopersmith <alan.coopersmith@oracle.com>
Reviewed-by: Alexandr Shadchin <alexandr.shadchin@gmail.com>
Reviewed-by: Peter Hutterer <peter.hutterer@who-t.net>
Error: Null pointer dereference (CWE 476)
Read from null pointer 'GetProtocol(proto)'
at line 3477 of src/mouse.c in function 'autoProbeMouse'.
Function 'GetProtocol' may return constant 'NULL' at line 736, called at line 3477.
Null pointer introduced at line 736 in function 'GetProtocol'.
We already handle one of the two cases that make GetProtocol return NULL,
proto == PROTO_UNKNOWN, but not PROT_UNSUP.
[ This bug was found by the Parfait 0.4.2 bug checking tool.
For more information see http://labs.oracle.com/projects/parfait/ ]
Signed-off-by: Alan Coopersmith <alan.coopersmith@oracle.com>
Reviewed-by: Alexandr Shadchin <alexandr.shadchin@gmail.com>
Fixes regression caused by commit b12fa0d5ab in which devices with
protocol "Auto" (generally just PS/2 devices on Solaris, since HAL
sets protocol "VUID" for USB devices already) weren't going through
vuidMousePreInit and setting up the device_control & read_input pointers
to the routines which know how to decode VUID packets.
Adds a check to vuidPreInit to ensure we don't leak memory if called
twice for the same device, just in case I missed a code path, or a
new one appears in the future.
Fixes Solaris bug 7070321: Mouse protocol "Auto" does not work in build 170
Signed-off-by: Alan Coopersmith <alan.coopersmith@oracle.com>
Acked-by: Peter Hutterer <peter.hutterer@who-t.net>
Appending NULL to an existing options list simply walks the entire
existing list before returning it unchanged, so if we aren't creating
a new list to merge, don't bother merging it either.
Signed-off-by: Alan Coopersmith <alan.coopersmith@oracle.com>
Reviewed-by: Peter Hutterer <peter.hutterer@who-t.net>
Commit 7bf22a36 (Use pInfo->options instead of conf-idev.) updated the
xf86CollectInputOptions call into keeping previous options (for ABI before
12). The hurd mouse driver also needs to be updated.
Fix regression introduced by 0a088df6: in the case of an OS-specific
protocol, the protocol is PROT_UNKNOWN, but should not be rejected: the core
mouse drive just needs to let the OS driver handle it.
Signed-off-by: Samuel Thibault <samuel.thibault@ens-lyon.org>
From Bug 34794:
"Until xf86-input-mouse-1.6.0, the MousePreInit() function exited if an
os specific PreInit function existed. It let the os-specific function finish
all initialisations that remained to be done after osInfo->PreInit() was
called.
The code in master now continues, and does things that the bsdMousePreInit()
function in bsd_mouse.c doesn't expect."
Fix up wsconsPreInit to deal with the new init process.
Fixes: https://bugs.freedesktop.org/show_bug.cgi?id=34794
Signed-off-by: Alexandr Shadchin <Alexandr.Shadchin@gmail.com>
Reviewed-by: Matthieu Herrb <matthieu.herrb@laas.fr>
Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net>
If it breaks, we know who to blame (or file a bug against).
Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net>
Reviewed-by: Cyril Brulebois <kibi@debian.org>
