]> sigrok.org Git - libsigrok.git/log
libsigrok.git
11 years agoDoxygen: Add @since markers.
Uwe Hermann [Sat, 13 Apr 2013 16:58:11 +0000 (18:58 +0200)]
Doxygen: Add @since markers.

Add @since tags to some (not all yet) public API functions, to document
in which release they were added (and/or in which release there were
API changes).

Document the @since usage in HACKING.

Also, add Doxygen docs for the functions in version.c.

11 years agoswig: wrap g_variant_new_uint64.
Martin Ling [Tue, 16 Apr 2013 02:33:56 +0000 (03:33 +0100)]
swig: wrap g_variant_new_uint64.

11 years agoswig: include cdata interface.
Martin Ling [Tue, 16 Apr 2013 02:11:48 +0000 (03:11 +0100)]
swig: include cdata interface.

11 years agoswig: add function to cast from void * to struct sr_datafeed_logic *
Martin Ling [Tue, 16 Apr 2013 02:02:52 +0000 (03:02 +0100)]
swig: add function to cast from void * to struct sr_datafeed_logic *

11 years agopython: support datafeed callbacks implemented in python.
Martin Ling [Tue, 16 Apr 2013 01:55:39 +0000 (02:55 +0100)]
python: support datafeed callbacks implemented in python.

11 years agoswig: use standard wrappers for integer types.
Martin Ling [Tue, 16 Apr 2013 01:55:11 +0000 (02:55 +0100)]
swig: use standard wrappers for integer types.

11 years agoAdd a void *cb_data parameter to datafeed callbacks.
Martin Ling [Mon, 15 Apr 2013 20:08:55 +0000 (21:08 +0100)]
Add a void *cb_data parameter to datafeed callbacks.

11 years agodrivers: return SR_ERR_NA on unsupported config key
Bert Vermeulen [Tue, 16 Apr 2013 00:33:03 +0000 (02:33 +0200)]
drivers: return SR_ERR_NA on unsupported config key

Fixes bug 89.

11 years agoPython bindings: Fix reported libsigrok version.
Uwe Hermann [Mon, 15 Apr 2013 22:58:29 +0000 (00:58 +0200)]
Python bindings: Fix reported libsigrok version.

Use 'pkg-config --modversion libsigrok' to get the libsigrok package
version. 'pkg-config --version libsigrok' ignores the argument and just
reports the version of pkg-config itself.

11 years agoPython bindings: Cosmetics.
Uwe Hermann [Mon, 15 Apr 2013 22:57:00 +0000 (00:57 +0200)]
Python bindings: Cosmetics.

11 years agoTell SWIG that gpointer is really void *.
Martin Ling [Mon, 15 Apr 2013 22:05:38 +0000 (23:05 +0100)]
Tell SWIG that gpointer is really void *.

11 years agoAdd SWIG interface file and build script for Python bindings.
Martin Ling [Mon, 15 Apr 2013 19:01:07 +0000 (20:01 +0100)]
Add SWIG interface file and build script for Python bindings.

11 years agools: Report when a device doesn't support metadata
Bert Vermeulen [Mon, 15 Apr 2013 22:23:34 +0000 (00:23 +0200)]
ols: Report when a device doesn't support metadata

Not an error, but it does mean it's either a really old or really
basic device, and will help us to debug people's problems with
these.

Fixes bug 92.

11 years agofx2lafw: Support conn scan parameter
Bert Vermeulen [Mon, 15 Apr 2013 21:50:16 +0000 (23:50 +0200)]
fx2lafw: Support conn scan parameter

This takes a USB specification. Fixes bug 82.

11 years agousb: Fix bus.address format detection
Bert Vermeulen [Mon, 15 Apr 2013 21:47:04 +0000 (23:47 +0200)]
usb: Fix bus.address format detection

11 years agools: Allow configuration before the device is opened
Bert Vermeulen [Mon, 15 Apr 2013 18:05:43 +0000 (20:05 +0200)]
ols: Allow configuration before the device is opened

11 years agorigol-ds1xx2: Don't assume there's a valid sdi
Bert Vermeulen [Mon, 15 Apr 2013 16:45:38 +0000 (18:45 +0200)]
rigol-ds1xx2: Don't assume there's a valid sdi

11 years agools: Properly initialize operational state before start
Bert Vermeulen [Mon, 15 Apr 2013 16:28:16 +0000 (18:28 +0200)]
ols: Properly initialize operational state before start

11 years agodemo: Fix compiler warning, cosmetics.
Uwe Hermann [Mon, 15 Apr 2013 10:27:55 +0000 (12:27 +0200)]
demo: Fix compiler warning, cosmetics.

11 years agolascar-el-usb: Use two probes for temp/humidity loggers
Bert Vermeulen [Sun, 14 Apr 2013 23:30:24 +0000 (01:30 +0200)]
lascar-el-usb: Use two probes for temp/humidity loggers

11 years agodemo: Unref the IO channel when it is no longer used
Joel Holdsworth [Sun, 14 Apr 2013 18:19:35 +0000 (19:19 +0100)]
demo: Unref the IO channel when it is no longer used

11 years agodemo: Free dev_contexts
Joel Holdsworth [Sun, 14 Apr 2013 17:52:02 +0000 (18:52 +0100)]
demo: Free dev_contexts

11 years agodemo: Moved demo configuration variables into dev_context
Joel Holdsworth [Sun, 14 Apr 2013 16:00:47 +0000 (17:00 +0100)]
demo: Moved demo configuration variables into dev_context

11 years agosession: Make sr_session_stop thread-safe
Alexandru Gagniuc [Sat, 2 Feb 2013 06:50:00 +0000 (00:50 -0600)]
session: Make sr_session_stop thread-safe

With the sigrok session running in a worker thread, if sr_session_stop is called
from another thread, it shuts down the pollfds used by the hardware drivers,
without ensuring that the sigrok event loop is no longer using those pollfds.

On the demo driver, this involves shutting down the GIOChannels, causing a
segfault when the sigrok event loop tries to use them. This is evident when
using the Stop button in PulseView, while the session is running.

This isn't a problem with just the demo driver; any driver's resources may be
freed by sr_session_stop concurrently with the sigrok session running.

To solve this problem, we don't touch the session itself in sr_session_stop().
Instead, we mark it for decommissioning and return. The session polls this flag,
and shuts itself down when requested.

This fixes bug 4.

Signed-off-by: Alexandru Gagniuc <redacted>
11 years agodemo: Keep a pointer to dev_context in sr_dev_inst and vice-versa
Joel Holdsworth [Sun, 14 Apr 2013 14:46:15 +0000 (15:46 +0100)]
demo: Keep a pointer to dev_context in sr_dev_inst and vice-versa

11 years agodemo: Only one GIOChannel is needed
Joel Holdsworth [Sat, 13 Apr 2013 15:43:16 +0000 (16:43 +0100)]
demo: Only one GIOChannel is needed

11 years agodemo: Fixed glitching in increment signal
Joel Holdsworth [Sat, 13 Apr 2013 15:08:52 +0000 (16:08 +0100)]
demo: Fixed glitching in increment signal

11 years agoalsa: Set the probe index
Joel Holdsworth [Sat, 16 Mar 2013 08:43:36 +0000 (08:43 +0000)]
alsa: Set the probe index

11 years agozeroplus-logic-cube: add USB ID for LAP-16128U
Jan Luebbe [Sat, 13 Apr 2013 21:27:07 +0000 (23:27 +0200)]
zeroplus-logic-cube: add USB ID for LAP-16128U

11 years agorigol-ds1xx2: support digital channels.
Martin Ling [Sun, 14 Apr 2013 01:21:55 +0000 (02:21 +0100)]
rigol-ds1xx2: support digital channels.

11 years agorigol-ds1xx2: doesn't actually support SR_CONF_LIMIT_SAMPLES
Martin Ling [Sat, 13 Apr 2013 21:25:41 +0000 (22:25 +0100)]
rigol-ds1xx2: doesn't actually support SR_CONF_LIMIT_SAMPLES

11 years agorigol-ds1xx2: fix handling of partial frames.
Martin Ling [Sun, 14 Apr 2013 00:58:35 +0000 (01:58 +0100)]
rigol-ds1xx2: fix handling of partial frames.

11 years agoignore test leftovers
Bert Vermeulen [Sun, 14 Apr 2013 14:44:49 +0000 (16:44 +0200)]
ignore test leftovers

11 years agools: Support get/set/list on all device options
Bert Vermeulen [Sat, 13 Apr 2013 22:36:05 +0000 (00:36 +0200)]
ols: Support get/set/list on all device options

Where it makes sense: _LOGIC_ANALYZER and _TRIGGER_TYPE don't have much
use for all of them.

11 years agorigol-ds1xx2: Use g_usleep() for portability
Bert Vermeulen [Fri, 12 Apr 2013 17:42:44 +0000 (19:42 +0200)]
rigol-ds1xx2: Use g_usleep() for portability

11 years agoBump glib requirement to >= 2.32.0.
Uwe Hermann [Fri, 12 Apr 2013 16:47:06 +0000 (18:47 +0200)]
Bump glib requirement to >= 2.32.0.

This is now required, e.g. due to the usage of g_variant_new_fixed_array().

11 years agorigol-ds1xx2: Whitespace, minor fix.
Uwe Hermann [Fri, 12 Apr 2013 16:44:28 +0000 (18:44 +0200)]
rigol-ds1xx2: Whitespace, minor fix.

Return SR_ERR_MALLOC for failed malloc, not SR_ERR.

11 years agounittests: Adapt to new GVariant system.
Uwe Hermann [Fri, 12 Apr 2013 15:59:38 +0000 (17:59 +0200)]
unittests: Adapt to new GVariant system.

11 years agoFix two minor compiler warnings.
Uwe Hermann [Fri, 12 Apr 2013 15:45:01 +0000 (17:45 +0200)]
Fix two minor compiler warnings.

11 years agooutput/vcd: Use selected probes, not lowest-numbered ones
Bert Vermeulen [Thu, 11 Apr 2013 16:49:24 +0000 (18:49 +0200)]
output/vcd: Use selected probes, not lowest-numbered ones

Thanks to Andrew Bardsley for the patch.

11 years agorigol-ds1xx2: Support for all channels, proper defaults
Bert Vermeulen [Thu, 11 Apr 2013 14:06:55 +0000 (16:06 +0200)]
rigol-ds1xx2: Support for all channels, proper defaults

Since this driver supports devices with a control panel and display,
we take the defaults from the device -- not a set of sensible
defaults as usual.

11 years agorigol-ds1xx2: fix channel numbers
Bert Vermeulen [Mon, 8 Apr 2013 10:01:00 +0000 (12:01 +0200)]
rigol-ds1xx2: fix channel numbers

Copyright bump, too.

11 years agorigol-ds1xx2: better debugging
Bert Vermeulen [Sun, 7 Apr 2013 23:12:42 +0000 (01:12 +0200)]
rigol-ds1xx2: better debugging

11 years agorigol-ds1xx2: properly send frame begin/end packets
Bert Vermeulen [Sun, 7 Apr 2013 22:38:58 +0000 (00:38 +0200)]
rigol-ds1xx2: properly send frame begin/end packets

11 years agohantek-dso: Support SR_CONF_NUM_TIMEBASE/SR_CONF_NUM_VDIV
Bert Vermeulen [Sun, 7 Apr 2013 22:37:50 +0000 (00:37 +0200)]
hantek-dso: Support SR_CONF_NUM_TIMEBASE/SR_CONF_NUM_VDIV

11 years agorigol-ds1xx2: Support SR_CONF_NUM_TIMEBASE/SR_CONF_NUM_VDIV
Bert Vermeulen [Sun, 7 Apr 2013 22:37:12 +0000 (00:37 +0200)]
rigol-ds1xx2: Support SR_CONF_NUM_TIMEBASE/SR_CONF_NUM_VDIV

11 years agoAdd SR_CONF keys to report the number of timebases and vdivs
Bert Vermeulen [Sun, 7 Apr 2013 22:29:37 +0000 (00:29 +0200)]
Add SR_CONF keys to report the number of timebases and vdivs

Number of timebases is needed to properly interpret a frame's worth
of samples. Number of vertical divisions isn't, but may nevertheless
be interesting for a frontend that wants to reproduce the scope's
native display.

11 years agouse driver callbacks directly
Bert Vermeulen [Sun, 7 Apr 2013 20:51:45 +0000 (22:51 +0200)]
use driver callbacks directly

11 years agorigol-ds1xx2: SR_CONF_TIMEBASE and _VDIVS lists are now an array of tuples
Bert Vermeulen [Sun, 7 Apr 2013 19:51:26 +0000 (21:51 +0200)]
rigol-ds1xx2: SR_CONF_TIMEBASE and _VDIVS lists are now an array of tuples

11 years agoreturn SR_ERR_ARG if a driver doesn't implement sr_config_*()
Bert Vermeulen [Sun, 7 Apr 2013 19:50:42 +0000 (21:50 +0200)]
return SR_ERR_ARG if a driver doesn't implement sr_config_*()

11 years agoAdd Rigol DS1052E/1102E VID:PID
Bert Vermeulen [Sun, 7 Apr 2013 19:28:54 +0000 (21:28 +0200)]
Add Rigol DS1052E/1102E VID:PID

11 years agohantek-dso: SR_CONF_TIMEBASE and _VDIVS lists are now an array of tuples
Bert Vermeulen [Sun, 7 Apr 2013 19:02:18 +0000 (21:02 +0200)]
hantek-dso: SR_CONF_TIMEBASE and _VDIVS lists are now an array of tuples

11 years agohantek-dso: move buffer sizes to device context
Bert Vermeulen [Tue, 2 Apr 2013 18:54:44 +0000 (20:54 +0200)]
hantek-dso: move buffer sizes to device context

There are only two sizes for every device: 10k and the complete
per-channel buffer size. This can be doubled if one channel is disabled.

11 years agools: Get rid of gcc warnings
Bert Vermeulen [Mon, 1 Apr 2013 14:53:14 +0000 (16:53 +0200)]
ols: Get rid of gcc warnings

11 years agochronovu-la8: Adjust to GVariant-based sr_config_* functions
Bert Vermeulen [Sun, 31 Mar 2013 19:31:49 +0000 (21:31 +0200)]
chronovu-la8: Adjust to GVariant-based sr_config_* functions

11 years agotondaj-sl-814: Adjust to GVariant-based sr_config_* functions
Bert Vermeulen [Sun, 31 Mar 2013 19:28:33 +0000 (21:28 +0200)]
tondaj-sl-814: Adjust to GVariant-based sr_config_* functions

11 years agouni-t-dmm: Adjust to GVariant-based sr_config_* functions
Bert Vermeulen [Sun, 31 Mar 2013 19:28:33 +0000 (21:28 +0200)]
uni-t-dmm: Adjust to GVariant-based sr_config_* functions

11 years agoserial-dmm: Adjust to GVariant-based sr_config_* functions
Bert Vermeulen [Sun, 31 Mar 2013 19:28:33 +0000 (21:28 +0200)]
serial-dmm: Adjust to GVariant-based sr_config_* functions

11 years agorigol-ds1xx2: Adjust to GVariant-based sr_config_* functions
Bert Vermeulen [Sun, 31 Mar 2013 19:25:51 +0000 (21:25 +0200)]
rigol-ds1xx2: Adjust to GVariant-based sr_config_* functions

11 years agomic-985xx: Adjust to GVariant-based sr_config_* functions
Bert Vermeulen [Sun, 31 Mar 2013 19:25:51 +0000 (21:25 +0200)]
mic-985xx: Adjust to GVariant-based sr_config_* functions

11 years agolink-mso19: Adjust to GVariant-based sr_config_* functions
Bert Vermeulen [Sun, 31 Mar 2013 19:25:51 +0000 (21:25 +0200)]
link-mso19: Adjust to GVariant-based sr_config_* functions

11 years agobrymen-dmm: Adjust to GVariant-based sr_config_* functions
Bert Vermeulen [Sun, 31 Mar 2013 18:55:39 +0000 (20:55 +0200)]
brymen-dmm: Adjust to GVariant-based sr_config_* functions

11 years agoasix-sigma: Adjust to GVariant-based sr_config_* functions
Bert Vermeulen [Sun, 31 Mar 2013 18:55:39 +0000 (20:55 +0200)]
asix-sigma: Adjust to GVariant-based sr_config_* functions

11 years agoalsa: Adjust to GVariant-based sr_config_* functions
Bert Vermeulen [Sun, 31 Mar 2013 18:55:39 +0000 (20:55 +0200)]
alsa: Adjust to GVariant-based sr_config_* functions

11 years agozeroplus-logic-cube: fix samplerate setting
Bert Vermeulen [Sun, 31 Mar 2013 18:47:01 +0000 (20:47 +0200)]
zeroplus-logic-cube: fix samplerate setting

11 years agosr_voltage_string(): deprecate struct sr_rational
Bert Vermeulen [Sun, 31 Mar 2013 08:27:15 +0000 (10:27 +0200)]
sr_voltage_string(): deprecate struct sr_rational

11 years agozeroplus-logic-cube: Adjust to GVariant-based sr_config_* functions
Bert Vermeulen [Sat, 30 Mar 2013 18:48:38 +0000 (19:48 +0100)]
zeroplus-logic-cube: Adjust to GVariant-based sr_config_* functions

11 years agonexus-osciprime: Adjust to GVariant-based sr_config_* functions
Bert Vermeulen [Sat, 30 Mar 2013 18:48:38 +0000 (19:48 +0100)]
nexus-osciprime: Adjust to GVariant-based sr_config_* functions

11 years agoagilent-dmm: Adjust to GVariant-based sr_config_* functions
Bert Vermeulen [Sat, 30 Mar 2013 18:48:38 +0000 (19:48 +0100)]
agilent-dmm: Adjust to GVariant-based sr_config_* functions

11 years agovictor-dmm: Adjust to GVariant-based sr_config_* functions
Bert Vermeulen [Sat, 30 Mar 2013 18:29:44 +0000 (19:29 +0100)]
victor-dmm: Adjust to GVariant-based sr_config_* functions

11 years agofluke-dmm: Adjust to GVariant-based sr_config_* functions
Bert Vermeulen [Sat, 30 Mar 2013 18:29:44 +0000 (19:29 +0100)]
fluke-dmm: Adjust to GVariant-based sr_config_* functions

11 years agocolead-slm: Adjust to GVariant-based sr_config_* functions
Bert Vermeulen [Sat, 30 Mar 2013 18:29:44 +0000 (19:29 +0100)]
colead-slm: Adjust to GVariant-based sr_config_* functions

11 years agolascar-el-usb: Adjust to GVariant-based sr_config_* functions
Bert Vermeulen [Sat, 30 Mar 2013 18:08:29 +0000 (19:08 +0100)]
lascar-el-usb: Adjust to GVariant-based sr_config_* functions

11 years agohantek-dso: Small debug fix
Bert Vermeulen [Sat, 30 Mar 2013 14:47:01 +0000 (15:47 +0100)]
hantek-dso: Small debug fix

11 years agohantek-dso: deprecate struct sr_rational
Bert Vermeulen [Sat, 30 Mar 2013 14:24:37 +0000 (15:24 +0100)]
hantek-dso: deprecate struct sr_rational

11 years agoinput/vcd: deprecate struct sr_rational
Bert Vermeulen [Sat, 30 Mar 2013 14:23:44 +0000 (15:23 +0100)]
input/vcd: deprecate struct sr_rational

11 years agosr_parse_(period|voltage): deprecate struct sr_rational
Bert Vermeulen [Sat, 30 Mar 2013 13:41:01 +0000 (14:41 +0100)]
sr_parse_(period|voltage): deprecate struct sr_rational

11 years agodemo: Adjust to GVariant-based sr_config_* functions
Bert Vermeulen [Tue, 26 Mar 2013 21:02:45 +0000 (22:02 +0100)]
demo: Adjust to GVariant-based sr_config_* functions

11 years agohantek-dso: Adjust to GVariant-based sr_config_* functions
Bert Vermeulen [Tue, 26 Mar 2013 20:38:32 +0000 (21:38 +0100)]
hantek-dso: Adjust to GVariant-based sr_config_* functions

11 years agostruct sr_samplerates is now obsolete
Bert Vermeulen [Mon, 25 Mar 2013 19:41:28 +0000 (20:41 +0100)]
struct sr_samplerates is now obsolete

sr_config_get() of SR_CONF_SAMPLERATE now returns a GVARIANT_TYPE_DICTIONARY.
This dictionary contains a single key-value pair. Possible values for the key:

 - "samplerates": the value is an array of GVARIANT_TYPE_UINT64 representing
   all valid samplerates.
 - "samplerate-steps": the value is an array of GVARIANT_TYPE_UINT64 with
   exactly three members, which represent the lowest samplerate, highest
   samplerate, and the minimum step, respectively.

11 years agools: allow for disabling of RLE as well
Bert Vermeulen [Mon, 25 Mar 2013 19:40:40 +0000 (20:40 +0100)]
ols: allow for disabling of RLE as well

11 years agools: Adjust to GVariant-based sr_config_* functions
Bert Vermeulen [Mon, 25 Mar 2013 19:40:15 +0000 (20:40 +0100)]
ols: Adjust to GVariant-based sr_config_* functions

11 years agofx2lafw: Adjust to GVariant-based sr_config_* functions
Bert Vermeulen [Mon, 25 Mar 2013 19:30:56 +0000 (20:30 +0100)]
fx2lafw: Adjust to GVariant-based sr_config_* functions

11 years agoAdd SR_CONF_TRIGGER_TYPE config_info list
Bert Vermeulen [Mon, 25 Mar 2013 19:29:45 +0000 (20:29 +0100)]
Add SR_CONF_TRIGGER_TYPE config_info list

11 years agoinput/output modules: Adjust to GVariant-based sr_config_* functions
Bert Vermeulen [Mon, 25 Mar 2013 19:27:26 +0000 (20:27 +0100)]
input/output modules: Adjust to GVariant-based sr_config_* functions

11 years agoAdd sr_config_free()
Bert Vermeulen [Mon, 25 Mar 2013 19:23:13 +0000 (20:23 +0100)]
Add sr_config_free()

Due to struct sr_config now containing a GVariant, this needs to be
cleaned up after use properly. This is a helper function for that.

11 years agoAdjust to GVariant-based sr_config_* functions
Bert Vermeulen [Mon, 25 Mar 2013 19:21:10 +0000 (20:21 +0100)]
Adjust to GVariant-based sr_config_* functions

11 years agoUse GVariant for sr_config_*() functions
Bert Vermeulen [Mon, 25 Mar 2013 14:38:44 +0000 (15:38 +0100)]
Use GVariant for sr_config_*() functions

sr_config_get() provides a GVariant owned by the caller, so it must be
released with g_variant_unref() when done.

sr_config_set() takes a GVariant from the caller which may be floating;
it will be properly sunk and release after use by this function. Thus
the output of g_variant_new_*() may be used as an argument.

sr_config_list() also provides a GVariant owned by the caller, to be
unreferenced when done.

sr_config_make() can take a floating reference.

11 years agoBump copyright year
Bert Vermeulen [Sun, 24 Mar 2013 10:21:00 +0000 (11:21 +0100)]
Bump copyright year

11 years agoconfigure.ac: Fix checks for cross-compiles.
Uwe Hermann [Tue, 9 Apr 2013 17:52:38 +0000 (19:52 +0200)]
configure.ac: Fix checks for cross-compiles.

When checking architecture-specific things, always check $host, i.e. the
architecture we're building _for_, not the one we happen to build _on_.

E.g. when cross-compiling _for_ Android (or Windows or others) it's important
to check for Android in $host; whether we happen to cross-compile _on_ a Linux
or Windows or OpenBSD or FreeBSD machine ($build) doesn't matter, only the
fact that we compile _for_ Android is important for most checks.

11 years agovcd output: Build fix for Windows/MinGW.
Uwe Hermann [Tue, 9 Apr 2013 16:50:34 +0000 (18:50 +0200)]
vcd output: Build fix for Windows/MinGW.

On Windows/MinGW 'recv' seems to be already defined in windows.h/winsock2.h.

Use 'receive' instead, for now, otherwise we get an error:

vcd.c:147:17: error: conflicting types for 'recv'

11 years agoconfigure.ac: Print $host and $build.
Uwe Hermann [Tue, 9 Apr 2013 16:47:31 +0000 (18:47 +0200)]
configure.ac: Print $host and $build.

In the configure summary at the end also print the architecture we're
building on ($build) and the target host we build for ($host). The two are
not necessarily the same, e.g. in the case of cross-compiles.

11 years agoconfigure.ac: Mention required lib versions.
Uwe Hermann [Tue, 9 Apr 2013 16:08:16 +0000 (18:08 +0200)]
configure.ac: Mention required lib versions.

In the summary output at the end of a configure run, explicitly mention
which versions of which libraries are required, and also the version which
pkg-config has found.

11 years agoconfigure.ac: Only enable rigol-ds1xx2 on Linux.
Uwe Hermann [Tue, 9 Apr 2013 15:58:37 +0000 (17:58 +0200)]
configure.ac: Only enable rigol-ds1xx2 on Linux.

11 years agoconfigure.ac: Cosmetics for the list of drivers.
Uwe Hermann [Tue, 9 Apr 2013 08:03:58 +0000 (10:03 +0200)]
configure.ac: Cosmetics for the list of drivers.

Use the canonical driver name (all-lowercase, e.g. "serial-dmm") in the
list of enabled/disabled drivers that configure prints after a run.

It's common to many drivers that they support multiple devices, so
printing one device name (e.g. "ChronoVu LA8") is seldom really correct.
E.g. the agilent-dmm, asix-sigma, brymen-dmm, colead-slm, fluke-dmm,
fx2lafw, hantek-dso, lascar-el-usb, mic-985xx, openbench-logic-sniffer,
rigol-ds1xx2, uni-t-dmm, victor-dmm, and zeroplus-logic-cube drivers
all support more than just one device.

So, just print the driver name instead which is more correct anyway
since it's specifically a list of enabled/disabled drivers.

11 years agoconfigure.ac: Improve a libusb-1.0 related check.
Uwe Hermann [Mon, 8 Apr 2013 13:56:54 +0000 (15:56 +0200)]
configure.ac: Improve a libusb-1.0 related check.

Don't rely on the "heuristic" that 'libusb_CFLAGS' will be non-empty if
libusb-1.0 was found, but rather use the proper method of checking the
variable 'have_libusb1_0' which pkg-config will set to "yes"/"no"
depending on whether it finds the library.

11 years agoconfigure.ac: Improve USB/EZUSB/serial checks.
Uwe Hermann [Mon, 8 Apr 2013 13:40:57 +0000 (15:40 +0200)]
configure.ac: Improve USB/EZUSB/serial checks.

 - If libusb-1.0 is not found, do not compile in ezusb.c and usb.c since
   they require libusb.h. The respective hardware drivers that use
   libusb-1.0, and usb.c and/or ezusb.c will be excluded from the build
   elsewhere in configure.ac. Rename NEED_EZUSB to NEED_USB.

 - Drop the NEED_SERIAL check and always compile in serial.c. This is a
   very small chunk of code, it does not depend on any external
   libraries that might be missing, and it compiles on all architectures.
   Thus there's no need to conditionally include or exclude it.

11 years agoconfigure.ac: Fix the order of some entries.
Uwe Hermann [Tue, 9 Apr 2013 10:32:30 +0000 (12:32 +0200)]
configure.ac: Fix the order of some entries.

In order for lib autodetection and disabling of drivers (which require
libs that cannot be found) to work, the order of AC_ARG_ENABLEs,
AM_CONDITIONALs, and AC_DEFINEs needs to be changed.

All drivers start out enabled or disabled (depending on whether
the --disable-drivers option was supplied or not). Then, any driver can
be enabled or disabled explicitly via --enable-<drivername> or the
resp. --disable-<drivername> option. Finally, pkg-config checks for all
libraries are performed, and all drivers which require a library that
cannot be found are disabled explicitly (regardless of any
--enable-<drivername> option that might have been supplied).

11 years agoconfigure.ac: Improve pkg-config related checks.
Uwe Hermann [Mon, 8 Apr 2013 12:46:13 +0000 (14:46 +0200)]
configure.ac: Improve pkg-config related checks.

Until now, we checked for certain (optional) libraries via pkg-config and
the configure script would abort if any of them was not found, even
though they were optional. It was up to the user to then figure out which
combination of --disable-<drivername> switches were required for his
specific OS (and set of installed libs) to get a working configure run.

Only if the user already specified enough --disable-<drivername>
switches beforehand, so that all drivers which require a missing library
were disabled, would the configure run not check for that specific lib
(and would thus not fail).

With this change, we now always unconditionally check for all libs
(required and optional) via pkg-config. However, whether an (optional) lib
is found or not, configure will not abort. Instead, it'll just disable
all drivers which need a lib that cannot be found.

The user will no longer have to supply --disable-<drivername> parameters
in order to get a working build.

11 years agoFix kernel driver attached check
Bert Vermeulen [Sun, 7 Apr 2013 12:14:16 +0000 (14:14 +0200)]
Fix kernel driver attached check

The check is really only broken on darwin, but useful on all other
platforms, even if only Linux can actually detach a driver.