linuxcnc.git
9 years agoWIP synchronized-homing
Chris Radek [Thu, 27 Jun 2013 00:13:32 +0000 (19:13 -0500)]
WIP

9 years agoread [AXIS*]HOME_SYNCHRONIZED and pass it along to motion
Chris Radek [Tue, 25 Jun 2013 18:29:31 +0000 (13:29 -0500)]
read [AXIS*]HOME_SYNCHRONIZED and pass it along to motion

9 years agoMake home-state a pin, to enable homing trickery
Chris Radek [Tue, 25 Jun 2013 17:20:39 +0000 (12:20 -0500)]
Make home-state a pin, to enable homing trickery

9 years agopyngcgui,gremlin_view apps and gladevcp widgets
Dewey Garrett [Sun, 24 Feb 2013 16:00:13 +0000 (09:00 -0700)]
pyngcgui,gremlin_view apps and gladevcp widgets

Python implementation of ngcgui functionality -- works with existing
ngcgui-compatible subroutines.  File selector in gui gives improved
hints for compatibility.  A numeric popup keyboard is available for
use with touchscreens.  Included:
  pyngcgui -- standalone application
  PyNgcGui -- gladevcp widget

Also:
  gremlin_view    -- standalone application for gremlin with common buttons
  HAL_GremlinPlus -- gladevcp widget

The gladevcp widgets can be embedded in guis like gscreen, touchy, gmoccapy

Example simulator ini files are provided for gscreen, touchy, gmoccapy
to show how to EMBED:

  configs/sim/ngcgui/pyngcgui_gscreen.ini
  configs/sim/ngcgui/pyngcgui_gscreen_popupkeyboard.ini

  configs/sim/ngcgui/pyngcgui_touchy.ini
  configs/sim/ngcgui/pyngcgui_touchy_lathe.ini

  configs/sim/ngcgui/pyngcgui_axis.ini

  configs/sim/gmoccapy/pyngcgui_gmoccapy.ini
  configs/sim/gmoccapy/pyngcgui_gmoccapy_with_popupkeyboard.ini

9 years agoDocs: fix inaccurate conflicting description
John Thornton [Mon, 24 Jun 2013 16:17:13 +0000 (11:17 -0500)]
Docs: fix inaccurate conflicting description

Signed-off-by: John Thornton <jthornton@gnipsel.com>
9 years agoDocs: fix inaccurate description
John Thornton [Sat, 22 Jun 2013 11:36:12 +0000 (06:36 -0500)]
Docs: fix inaccurate description

Signed-off-by: John Thornton <jthornton@gnipsel.com>
9 years agoedge: fix spurious edge; add new trigger mode
Jeff Epler [Sat, 22 Jun 2013 03:24:09 +0000 (22:24 -0500)]
edge: fix spurious edge; add new trigger mode

John Kasunich noted that edge would have an output pulse when the
in-edge parameter was modified.  Fix this bug.

He was looking because he wanted a trigger on both edges; add this
at the same time.  Happily, it turns out to be cleaner with the new
desired_edge logic.

9 years agoMerge branch 'v2.5_branch'
Sebastian Kuzminsky [Fri, 21 Jun 2013 22:33:45 +0000 (17:33 -0500)]
Merge branch 'v2.5_branch'

9 years agofix pyvcp dial widget behavior at min/max limit
John Kasunich [Fri, 21 Jun 2013 22:20:02 +0000 (17:20 -0500)]
fix pyvcp dial widget behavior at min/max limit

the overflow/underflow detection logic misbehaved
for some ratios of limit and increment size, result
was a value that wrapped instead of clamping at the
limit

9 years agomake loadusr search the bin/ subdir of the configuration
Jeff Epler [Fri, 21 Jun 2013 21:21:02 +0000 (16:21 -0500)]
make loadusr search the bin/ subdir of the configuration

9 years agocompensate for lag due to servo period
John Kasunich [Fri, 21 Jun 2013 03:55:41 +0000 (22:55 -0500)]
compensate for lag due to servo period

9 years agohalmodule: fix deallocation of hal objects
Jeff Epler [Fri, 21 Jun 2013 03:43:49 +0000 (22:43 -0500)]
halmodule: fix deallocation of hal objects

There's no guarantee that PyObject_Del is the right underlying
function to call; use the one from our type object instead.  Without
this, sometimes a crash was observed at shutdown.

9 years agortai_ulapi: ensure that rtapi's internal data is unmapped when appropriate
Jeff Epler [Fri, 21 Jun 2013 03:19:34 +0000 (22:19 -0500)]
rtai_ulapi: ensure that rtapi's internal data is unmapped when appropriate

this helps support cmorley's use case of having a single process which
lives through multiple realtime sessions, creating a realtime component
in each one.

9 years agonml cleanup: make keystick identify itself as 'xemc'
Sebastian Kuzminsky [Thu, 20 Jun 2013 17:17:50 +0000 (12:17 -0500)]
nml cleanup: make keystick identify itself as 'xemc'

keystick doesn't need its own NML access id, just use the 'xemc' id like
all the other UIs do.

9 years agonml cleanup: the GUIs don't need access to the tool NML channels
Sebastian Kuzminsky [Thu, 20 Jun 2013 17:10:16 +0000 (12:10 -0500)]
nml cleanup: the GUIs don't need access to the tool NML channels

The tool command channel 'toolCmd' and the tool status channel 'toolSts'
are only used between Task and IO.  The GUIs don't need access to them.

9 years agonml cleanup: IO never writes to the toolCmd channel
Sebastian Kuzminsky [Wed, 19 Jun 2013 23:02:12 +0000 (17:02 -0600)]
nml cleanup: IO never writes to the toolCmd channel

IO reads from the toolCmd channel but doesn't write to it, so remove
its write permissions from the NML file.

9 years agohalmodule: produce python exceptions instead of segfaults
Jeff Epler [Thu, 20 Jun 2013 23:31:05 +0000 (18:31 -0500)]
halmodule: produce python exceptions instead of segfaults

.. in the case of incorrect use of most attributes of hal.component
instances after they've been exited via the .exit() method.

9 years agotestsuite: rehabilitate mux test
Jeff Epler [Thu, 20 Jun 2013 16:22:13 +0000 (11:22 -0500)]
testsuite: rehabilitate mux test

make everything run in a single thread so that it is not an
unintentional test of thread interleaving semantics.  Update the
result because now some output transitions consistently happen one
servo cycle earlier than before.

The test now passes (200x) on my machine with PREEMPTY_RT when
merged into the rtos branch.  On the same test machine, the old test
failed somewhere between 1-in-10 and 1-in-50 of the time.

9 years agoturn on output from all "unreliable" tests and unskip
Sebastian Kuzminsky [Fri, 14 Jun 2013 21:41:33 +0000 (15:41 -0600)]
turn on output from all "unreliable" tests and unskip

9 years agoreenable t0 tests
Chris Radek [Wed, 29 May 2013 21:03:36 +0000 (16:03 -0500)]
reenable t0 tests

9 years agohalui-jogging test: let the test script choose pass/fail
Sebastian Kuzminsky [Fri, 14 Jun 2013 21:24:05 +0000 (15:24 -0600)]
halui-jogging test: let the test script choose pass/fail

This commit makes test success independent of output, since there is
some ambiguity in test output (unexpected realtime delays, 32/64-bit
float rounding errors, etc).

If test.sh fails, the test fails (just like before).  If test.sh succeeds,
the test passes no matter what the output was.

9 years agotests: make halui-jogging report its problems
Chris Radek [Wed, 29 May 2013 19:29:56 +0000 (14:29 -0500)]
tests: make halui-jogging report its problems

Also turn off this test's debug output since it's not portable between
i386 and amd64.  Fortunately the output of the "ui" is all we really need.

9 years agolinuxcnc script: add -r for runtests
Chris Radek [Wed, 29 May 2013 19:21:56 +0000 (14:21 -0500)]
linuxcnc script: add -r for runtests

This tells the script to not mess with stdout and stderr.  This is
important for runtests that use a custom "ui" to exercise something.

9 years agobuild: fix dependency analyis after changing branches
Jeff Epler [Wed, 19 Jun 2013 13:06:01 +0000 (08:06 -0500)]
build: fix dependency analyis after changing branches

this fixes incomplete dependency analysis after switching from
v2.5_branch to master branch, the overt symptom of which was a failing
build with an error like this:
emcIniFile.cc:66: undefined reference to
 `IniFile::Find(int*, IniFile::StrIntPair*, char const*, char const*, int)'

fix this by reading the dependency information from where v2.5_branch
would have left it if there's not one where we would have created it.

9 years agoladder: remove stray # added by mistake
Jeff Epler [Mon, 17 Jun 2013 19:23:41 +0000 (14:23 -0500)]
ladder: remove stray # added by mistake

9 years agoMerge remote-tracking branch 'origin/v2.5_branch'
Jeff Epler [Mon, 17 Jun 2013 18:58:52 +0000 (13:58 -0500)]
Merge remote-tracking branch 'origin/v2.5_branch'

9 years agoDelete unused file
Jeff Epler [Mon, 17 Jun 2013 15:50:47 +0000 (10:50 -0500)]
Delete unused file

9 years agoAdd license notices where they were missing
Jeff Epler [Mon, 17 Jun 2013 15:50:40 +0000 (10:50 -0500)]
Add license notices where they were missing

9 years agoMerge remote-tracking branch 'origin/v2.5_branch'
Sebastian Kuzminsky [Thu, 13 Jun 2013 21:50:48 +0000 (15:50 -0600)]
Merge remote-tracking branch 'origin/v2.5_branch'

9 years agodoc: asciidoc doesnt understand translated paragraph styles
Sebastian Kuzminsky [Thu, 13 Jun 2013 20:34:49 +0000 (14:34 -0600)]
doc: asciidoc doesnt understand translated paragraph styles

9 years agodocs: fix an asciidoc markup error in m-code docs
Sebastian Kuzminsky [Thu, 13 Jun 2013 19:22:06 +0000 (13:22 -0600)]
docs: fix an asciidoc markup error in m-code docs

9 years agodocs: fix a syntax error typo in gladevcp manpage
Sebastian Kuzminsky [Thu, 13 Jun 2013 18:43:39 +0000 (12:43 -0600)]
docs: fix a syntax error typo in gladevcp manpage

9 years agogmoccapy - locale included ! Solved some smaler isues
Norbert Schechner [Tue, 11 Jun 2013 17:27:16 +0000 (19:27 +0200)]
gmoccapy - locale included ! Solved some smaler isues

ver. 0.9.4
- corrected a lot of typo, thanks to David for his help
- renamed some widgets from "cero" to "zero"
- deleted one increment in every INI file, so the GUI fits better to smalest screen size
- translation to german included and prepared for more langauges
- there was still an error in key handling with incremental jogging, hope I solved it now for final
- changed the dialog "run_from_line" to use gscreens default, resulting in the need to rename
  "hal_sourceview" to "gcode_view"
- renamed all pins and definitions from "overide" to "override" because of typo
- avoid getting to the touch off button, while the machine is not homed
- hide Y touch off buttons in lathe mode
- solved a bug executing a command constantly related with feed and periodic
- Tool change from tool editor now works like it should even with T1 M6 as with M61 Q1
- the tool in spindle will be checkt if you enter the tool editor.

Signed-off-by: Chris Morley <chrisinnanaimo@hotmail.com>
9 years agodisable mux test until I get time to fix it
Andy Pugh [Mon, 10 Jun 2013 21:56:57 +0000 (22:56 +0100)]
disable mux test until I get time to fix it

Signed-off-by: Andy Pugh <andy@bodgesoc.org>
9 years agoadd velocity estimation via encoder timestamp to PPMC encoder
elson [Mon, 10 Jun 2013 03:27:19 +0000 (22:27 -0500)]
add velocity estimation via encoder timestamp to PPMC encoder

9 years agogladevcp -add a set_selected_tool method to tooledit_widget
Chris Morley [Sun, 9 Jun 2013 19:21:03 +0000 (12:21 -0700)]
gladevcp -add a set_selected_tool method to tooledit_widget

highlights and checks the specified toolnumber.

9 years agogladevcp -calculator widget: replace tabs with spaces
Chris Morley [Sun, 9 Jun 2013 16:59:08 +0000 (09:59 -0700)]
gladevcp -calculator widget: replace tabs with spaces

no other changes intended. for my sanity.

9 years agogladevcp fix calculator widget math error
Chris Morley [Sun, 9 Jun 2013 16:43:17 +0000 (09:43 -0700)]
gladevcp fix calculator widget math error

numbers starting with a 0 such as 0123+1 would be
converted to octal instead of decimal (so above = 84)
also dividing a whole number into a faction would not
show the decimal portion (1/2 would not = .5)
now we split the string up and qualify all the parts.
converting the numbers to floats first. Then we put it back
together and eval() it.

This bug affected Gscreen's run-from-line badly and certain
division calculations as above.

9 years agoMerge branch 'v2.5_branch'
Chris Morley [Sun, 9 Jun 2013 05:13:22 +0000 (22:13 -0700)]
Merge branch 'v2.5_branch'

9 years agopncconf -fix an open loop test error.
Chris Morley [Sun, 9 Jun 2013 04:32:43 +0000 (21:32 -0700)]
pncconf -fix an open loop test error.

setting GPIO to output would cause a HAL error.
If inverting PWM ther would be a HAL error.

9 years agoremove dead code
Sebastian Kuzminsky [Fri, 7 Jun 2013 17:50:46 +0000 (11:50 -0600)]
remove dead code

This helper is no longer needed, since moving to asciidoc.

9 years agofix an "uninitialized variable" warning
Sebastian Kuzminsky [Wed, 5 Jun 2013 03:09:57 +0000 (21:09 -0600)]
fix an "uninitialized variable" warning

9 years agogs2: simplify --verbose/--debug
Sebastian Kuzminsky [Wed, 5 Jun 2013 02:58:00 +0000 (20:58 -0600)]
gs2: simplify --verbose/--debug

The code makes no distinction between --debug and --verbose, so
squash those arguments together.  Leave both for now, for backwards
compatibility.

Also fixes a compiler warning about the variable "verbose" being set
but not used.

9 years agoremove the old forked modbus code
Sebastian Kuzminsky [Wed, 22 May 2013 17:50:19 +0000 (11:50 -0600)]
remove the old forked modbus code

Now that gs2_vfd is updated to build against libmodbus3 (supplied by
official debian packages for Precise, and by linuxcnc backported debian
packages for Hardy and Lucid) there are no users of our forked old copy
of modbus, so remove it.

9 years agoswitch gs2_vfd to use libmodbus3
Sebastian Kuzminsky [Wed, 22 May 2013 17:50:35 +0000 (11:50 -0600)]
switch gs2_vfd to use libmodbus3

Before this commit, gs2_vfd used our old forked/hacked copy of libmodbus2.

This commit changes gs2_vfd to use the new libmodbus3 API instead.
The change to the C code is largely mechanical.

Also some tweaks to the makefile:

    Don't build gs2 if the user configured with --without-libmodbus.

    Use the libmodbus CFLAGS and LIBS that configure determined.

9 years agoconfigure: require libmodbus3
Sebastian Kuzminsky [Thu, 30 May 2013 15:44:37 +0000 (09:44 -0600)]
configure: require libmodbus3

By default, configure will now look for libmodbus (installed from a
separate deb, provided by Ubuntu for Precise and by the linuxcnc.org
debian archive for Hardy and Lucid) and error out if it can't be found.

The user can disable this check (and disable compilation of all drivers
that use libmodbus) by configuring with "--without-libmodbus".

9 years agobuild-depend on libmodbus-dev (3.0 or newer)
Sebastian Kuzminsky [Wed, 22 May 2013 15:21:29 +0000 (09:21 -0600)]
build-depend on libmodbus-dev (3.0 or newer)

Currently, linuxcnc uses three different modbus implementations:

    gs2_vfd uses a local copy of libmodbus2 (forked into the linuxcnc
    repo in 2008, hacked with bug fixes over the years)

    vfs11_vfd and mb2hal use libmodbus3, but it was until recently not
    available on Hardy or Lucid (see below)

    classicladded uses its own internal modbus implementation

I took the libmodbus3 source package from Precise, applied some
tiny tweaks to get it to build on Hardy and Lucid, and built it for
those platforms.  The resulting binary debs are now available in the
linuxcnc.org debian archive.

This commit changes linuxcnc to build-depend on the libmodbus package.
It uses the upstream libmodbus3 on Precise, and these new local backported
packages on Hardy and Lucid.  It looks to me like all our local bugfixes
have gone into libmodbus3 (the newly available packaged version).

9 years agoMerge branch 'v2.5_branch'
Sebastian Kuzminsky [Wed, 5 Jun 2013 03:04:07 +0000 (21:04 -0600)]
Merge branch 'v2.5_branch'

9 years agodocs: fix mistakes in the gs2 manpage
Sebastian Kuzminsky [Wed, 5 Jun 2013 03:02:49 +0000 (21:02 -0600)]
docs: fix mistakes in the gs2 manpage

The long arguments need two dashes, not one.

9 years agoRevert part of "gmoccapy 0.9.2 - now with lathe support"
Chris Radek [Tue, 4 Jun 2013 14:35:17 +0000 (09:35 -0500)]
Revert part of "gmoccapy 0.9.2 - now with lathe support"

This reverts an accidental sim.tbl change in commit
23ba9827a3c31cdb33b51e3d33c26985745fa64d
pointed out by Dewey Garrett on emc-developers.

9 years agoDocs: add info about thread requirements
John Thornton [Tue, 4 Jun 2013 11:54:34 +0000 (06:54 -0500)]
Docs: add info about thread requirements

Signed-off-by: John Thornton <jthornton@gnipsel.com>
9 years agoDocs: fix typo option number
John Thornton [Mon, 3 Jun 2013 20:03:35 +0000 (15:03 -0500)]
Docs: fix typo option number

Signed-off-by: John Thornton <jthornton@gnipsel.com>
9 years agogscreen -allow custom skins to supply a locale tranlation.
Chris Morley [Fri, 31 May 2013 06:02:13 +0000 (23:02 -0700)]
gscreen -allow custom skins to supply a locale tranlation.

for Norbert's Gmoccapy

Signed-off-by: Chris Morley <chrisinnanaimo@hotmail.com>
9 years agoPut back the mux test, hopefully fixed
Andy Pugh [Fri, 31 May 2013 18:11:31 +0000 (19:11 +0100)]
Put back the mux test, hopefully fixed
(Hope that pushing the input data helps)

Signed-off-by: Andy Pugh <andy@bodgesoc.org>
9 years agodocs: fix manpage markup for mux_generic
Sebastian Kuzminsky [Fri, 31 May 2013 06:16:15 +0000 (00:16 -0600)]
docs: fix manpage markup for mux_generic

9 years agogmoccapy - Z axis had wrong directions by keyboard jogging and introduced a check...
Norbert Schechner [Tue, 28 May 2013 07:51:16 +0000 (09:51 +0200)]
gmoccapy - Z axis had wrong directions by keyboard jogging and introduced a check to avoid spindle speed over maximum given

ver. 0.9.3.2
- alligned the jog_button and jog_rates frame on the top (just cosmetic)
- The Page-up and Page_Down keyboard keys moved the Z axis in oposide directions
- Feed values will only be shown with digit if G95 is active
- added classicladder button to the list of buttons to be able to handle it with hardware button
- changed hardware button handling to fit also to lathe mode
- included a test to avoid speeding the spindle over its limit using the spindle speed overide
  i.e. max = 6000 but S = 5500 and overide = 120 % would result in 6600, but only 6000 is possible
- also check spindle speed if the commands are given trough MDI and overide is to high will reduce
  the overide value to max allowed

Signed-off-by: Chris Morley <chrisinnanaimo@hotmail.com>
9 years agoDisable the mux test to get the buildbot working again
Andy Pugh [Fri, 31 May 2013 00:48:10 +0000 (01:48 +0100)]
Disable the mux test to get the buildbot working again

Signed-off-by: Andy Pugh <andy@bodgesoc.org>
9 years agoAdd a new generic multiplexer (mux) component
Andy Pugh [Thu, 30 May 2013 21:13:43 +0000 (22:13 +0100)]
Add a new generic multiplexer (mux) component
Any number of pins, any data type.

Signed-off-by: Andy Pugh <andy@bodgesoc.org>
9 years agoMerge branch 'v2.5_branch'
Sebastian Kuzminsky [Thu, 30 May 2013 19:56:41 +0000 (13:56 -0600)]
Merge branch 'v2.5_branch'

9 years agofix a cut-n-paste error in configure.in
Sebastian Kuzminsky [Thu, 30 May 2013 15:43:57 +0000 (09:43 -0600)]
fix a cut-n-paste error in configure.in

9 years agoDocs: remove trailing whitespace
John Thornton [Thu, 30 May 2013 11:44:45 +0000 (06:44 -0500)]
Docs: remove trailing whitespace

Signed-off-by: John Thornton <jthornton@gnipsel.com>
9 years agoDocs: fix markup error
John Thornton [Thu, 30 May 2013 11:43:14 +0000 (06:43 -0500)]
Docs: fix markup error

Signed-off-by: John Thornton <jthornton@gnipsel.com>
9 years agoDocs: fix incorrect pin name in example text
John Thornton [Wed, 29 May 2013 19:48:04 +0000 (14:48 -0500)]
Docs: fix incorrect pin name in example text

Signed-off-by: John Thornton <jthornton@gnipsel.com>
9 years agoDocs: add info on homing a rotary axis
John Thornton [Wed, 29 May 2013 12:41:08 +0000 (07:41 -0500)]
Docs: add info on homing a rotary axis

Signed-off-by: John Thornton <jthornton@gnipsel.com>
9 years agohal/vfs_vfd11: fix missing pin pointer deref
Michael Haberler [Tue, 28 May 2013 17:48:00 +0000 (19:48 +0200)]
hal/vfs_vfd11: fix missing pin pointer deref

The original code would never update "less urgent registers"
even when (pullcount == 0).

thanks to Yishin Li!

9 years agogmoccapy - changed screen size requierment and warning from G95 bug
Norbert Schechner [Mon, 27 May 2013 17:45:23 +0000 (19:45 +0200)]
gmoccapy - changed screen size requierment and warning from G95 bug

ver. 0.9.3.1
- A message window will warn the user in case he try to use G95
- Changed the requiered screen size to fit on 979 x 750

Signed-off-by: Chris Morley <chrisinnanaimo@hotmail.com>
9 years agogmoccapy - fixed a bug in max velocity handling and added patches from gscreen
Norbert Schechner [Wed, 22 May 2013 19:38:17 +0000 (21:38 +0200)]
gmoccapy - fixed a bug in max velocity handling and added patches from gscreen

ver. 0.9.3
- included the patch for key_jogging from Chris
- support now also back tool lathes
- after M61 Q tool change, the Gui didn't leave the MDI modem resulting in problems
- the btn_view_Y2 has now a yellow background
- changing the max vel slider had no efekt, I fixed this

Signed-off-by: Chris Morley <chrisinnanaimo@hotmail.com>
9 years agoMerge branch 'v2.5_branch'
Sebastian Kuzminsky [Wed, 22 May 2013 17:52:44 +0000 (11:52 -0600)]
Merge branch 'v2.5_branch'

9 years agodocs: fix whitespace in the m66 section
Sebastian Kuzminsky [Wed, 22 May 2013 17:52:32 +0000 (11:52 -0600)]
docs: fix whitespace in the m66 section

9 years agodocs: fix a troff markup typo in gs2 manpage
Sebastian Kuzminsky [Tue, 21 May 2013 21:55:39 +0000 (15:55 -0600)]
docs: fix a troff markup typo in gs2 manpage

9 years agoDocs: re-branding screenshots
John Thornton [Mon, 20 May 2013 23:34:45 +0000 (18:34 -0500)]
Docs: re-branding screenshots

Signed-off-by: John Thornton <jthornton@gnipsel.com>
9 years agogscreen configs -add industrial lathe sample config
Chris Morley [Mon, 20 May 2013 03:59:21 +0000 (20:59 -0700)]
gscreen configs -add industrial lathe sample config

9 years agogscreen config -fix diameter mode to actually change
Chris Morley [Mon, 20 May 2013 03:55:19 +0000 (20:55 -0700)]
gscreen config -fix diameter mode to actually change

industrial uses the dro widget so must have a function called
to change the mode

9 years agogladevcp -fix drowidget radius mode function
Chris Morley [Mon, 20 May 2013 03:53:27 +0000 (20:53 -0700)]
gladevcp -fix drowidget radius mode function

It was miss named

9 years agocomp: export() doesnt need buf if there are no functions
Sebastian Kuzminsky [Sun, 19 May 2013 23:49:42 +0000 (17:49 -0600)]
comp: export() doesnt need buf if there are no functions

This commit removes an "unused variable" compiler warning when compiling
comps without functions, such as mesa_7i65.

9 years agodocs: move linuxcnc.1.in out of the manpage tree
Sebastian Kuzminsky [Fri, 17 May 2013 17:57:02 +0000 (13:57 -0400)]
docs: move linuxcnc.1.in out of the manpage tree

The man/man? directories are for manpages, not for the source code
for manpages.

mandb complains about any non-manpage files it finds in there.

We have one non-manpage file in those directories: man1/linuxcnc.1.in,
which is turned into linuxcnc.1 by config.status (called by make),
to embed the search path for .ini files into the manpage.

This commit moves linuxcnc.1.in into the docs/src tree, so as not to
pollute the manpage directory tree.

9 years agodont build linuxcnc.1 from configure, it gets built by make
Sebastian Kuzminsky [Fri, 17 May 2013 21:01:10 +0000 (17:01 -0400)]
dont build linuxcnc.1 from configure, it gets built by make

9 years agoMerge branch 'v2.5_branch'
Sebastian Kuzminsky [Sun, 19 May 2013 16:09:39 +0000 (12:09 -0400)]
Merge branch 'v2.5_branch'

9 years agorebranding
Sebastian Kuzminsky [Sun, 19 May 2013 16:08:45 +0000 (12:08 -0400)]
rebranding

9 years agorebranding fixes
Sebastian Kuzminsky [Sun, 19 May 2013 16:05:39 +0000 (12:05 -0400)]
rebranding fixes

9 years agothe 'mdi' script no longer needs an nml file
Sebastian Kuzminsky [Sun, 19 May 2013 15:53:51 +0000 (11:53 -0400)]
the 'mdi' script no longer needs an nml file

It can use the default one, just like every other UI.

9 years agorebranding
Sebastian Kuzminsky [Sun, 19 May 2013 15:44:35 +0000 (11:44 -0400)]
rebranding

9 years agodocs: fix an asciidoc markup syntax error
Sebastian Kuzminsky [Sun, 19 May 2013 15:27:29 +0000 (11:27 -0400)]
docs: fix an asciidoc markup syntax error

9 years agorebranding
Sebastian Kuzminsky [Sun, 19 May 2013 15:27:10 +0000 (11:27 -0400)]
rebranding

9 years agoMerge branch 'v2.5_branch'
Sebastian Kuzminsky [Sun, 19 May 2013 02:22:15 +0000 (22:22 -0400)]
Merge branch 'v2.5_branch'

9 years agorebranding in the output of configure
Sebastian Kuzminsky [Sun, 19 May 2013 00:46:48 +0000 (20:46 -0400)]
rebranding in the output of configure

9 years agoMerge branch 'v2.5_branch'
Sebastian Kuzminsky [Sat, 18 May 2013 20:48:05 +0000 (14:48 -0600)]
Merge branch 'v2.5_branch'

9 years agoshebang python, not python2
Sebastian Kuzminsky [Sat, 18 May 2013 20:42:41 +0000 (14:42 -0600)]
shebang python, not python2

All the platforms i've checked (Hardy, Lucid, Precise, and Wheezy) have
a symlink named python pointing at a python2 interpreter, but not all
have a symlink named python2.

If/when python3 becomes widely used we'll have to solve this in a more
clever way, but for now it works fine.

9 years agofix "runtests -v"
Sebastian Kuzminsky [Sat, 18 May 2013 20:40:37 +0000 (14:40 -0600)]
fix "runtests -v"

9 years agodon't require bash when any posix sh will do
Sebastian Kuzminsky [Sat, 18 May 2013 19:47:12 +0000 (13:47 -0600)]
don't require bash when any posix sh will do

9 years agodocs: fix a typo in the halcmd manpage
Sebastian Kuzminsky [Fri, 17 May 2013 13:42:39 +0000 (09:42 -0400)]
docs: fix a typo in the halcmd manpage

9 years agocomp: clean up generated C code
Sebastian Kuzminsky [Fri, 17 May 2013 01:58:46 +0000 (21:58 -0400)]
comp: clean up generated C code

If the .comp has no extra_setup() function that might modify
inst->_personality, don't bother resetting the 'personality' local
variable.

9 years agoNULL is not an int, but 0 is
Sebastian Kuzminsky [Thu, 16 May 2013 19:44:57 +0000 (15:44 -0400)]
NULL is not an int, but 0 is

This shuts up some "-Wpointer-arith" compiler warnings:

    warning: NULL used in arithmetic

9 years agogenserkins: reduce stack usage
Sebastian Kuzminsky [Thu, 16 May 2013 18:01:45 +0000 (14:01 -0400)]
genserkins: reduce stack usage

Normally I like limiting the scope of local variables as much as
possible, but in this case two adjacent scopes both instanciate a large
anonymous struct, and the compiler doesn't realize they're identical
and have non-overlapping use, and it puts both of them on the stack.
This results in a very large stack frame for this function.

This commit moves the declaration of the large structure out into the
enclosing scope, and both of the two nested scopes use the shared struct.

9 years agoun-ignore some files we no longer have
Sebastian Kuzminsky [Thu, 16 May 2013 15:38:06 +0000 (11:38 -0400)]
un-ignore some files we no longer have

9 years agobuild system: show mandb command if the user requested it
Sebastian Kuzminsky [Thu, 16 May 2013 15:20:57 +0000 (11:20 -0400)]
build system: show mandb command if the user requested it

9 years agoignore files & dirs generated by mandb
Sebastian Kuzminsky [Thu, 16 May 2013 15:19:50 +0000 (11:19 -0400)]
ignore files & dirs generated by mandb

9 years agoignore new generated manpages
Sebastian Kuzminsky [Thu, 16 May 2013 15:11:58 +0000 (11:11 -0400)]
ignore new generated manpages

9 years agosort & remove duplicates
Sebastian Kuzminsky [Thu, 16 May 2013 15:11:27 +0000 (11:11 -0400)]
sort & remove duplicates

This tidies up the gitignore info for manpages.  Reordering & removing
dupes only, no additions or subtractions.

9 years agodocs: reorg gitignore to placate mandb
Sebastian Kuzminsky [Thu, 16 May 2013 15:09:26 +0000 (11:09 -0400)]
docs: reorg gitignore to placate mandb

We run mandb to index all our manpages.  mandb complains about every
non-manpage file it finds in the man directory tree.  This includes our
.gitignore files.

This commit collects all the .gitignore info into the .gitignore file
in docs/man, just outside of where mandb starts paying attention and
complaining.