summaryrefslogtreecommitdiff
path: root/xts5
AgeCommit message (Collapse)AuthorFilesLines
2010-06-28Merge remote branch 'whot/master'HEADmasterDan Nicholson5-60/+130
Conflicts: xts5/Makefile.am Moved the DISPLAY check to the top-level Makefile.am where check-local is now defined.
2010-06-25Xlib12: Add GenericEvent to list of permitted events.Peter Hutterer2-2/+3
Fixes error message "_xgeEventToWire: Unknown extension -1, this should never happen." on Xlib12 XSendEvent 21. This requires expanding the global event_type array to LASTEvent - 2 instead of the current hardcoded 32. Since the first two opcodes are used for replies and errors, two minus the last event opcode is sufficient. Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net>
2010-06-25Fail if DISPLAY is unset.Peter Hutterer1-0/+5
All tests require DISPLAY to be set and it is rather annoying when a test runs four hours just to have all the results complain about a missing DISPLAY environment variable. Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net>
2010-06-25Xlib13 XGrabButton: don't overrun the number of buttons configured.Peter Hutterer1-2/+3
NS_button contains the static button mappings (for whatever reason). Don't try to access it nbuttons() time because that will return the number of buttons configured on the device (usually 12). Instead, use the size of the array we're traversing through. Aside from running OOB on an array being considered impolite, it also gives us BadValue when trying to XTestFakeButtonEvent a Button 0, which the test suite doesn't like. Fixes UNRESOLVED on test case 10/39. Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net>
2010-06-25Add -h help output to pt.Peter Hutterer1-11/+18
Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net>
2010-06-24Assign a couple of default values to the config.Peter Hutterer1-1/+14
Since most values are optional, set a couple of reasonable defaults, thus making the need for TET_CONFIG obsolete. Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net>
2010-06-24Make all config variables optional.Peter Hutterer1-35/+36
Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net>
2010-06-24Retrieve config variables from the environment.Peter Hutterer1-2/+37
After parsing the config file, run again through the environment and check if any of the variables (notably DISPLAY, added here) was set there. If so, overwrite the config value. Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net>
2010-06-24Add -gdb option to pt script.Peter Hutterer1-1/+6
pt sets up some of the environment vital to the execution of the test. Add a -gdb flag that simply starts the program through gdb with that enviroment. The user just has to hit "run" in gdb to debug. Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net>
2010-06-24Improve pt script to allow for single test-cases to be run.Peter Hutterer1-7/+14
This patch adds two changes: Instead of defaulting to a ./Test binary (these aren't built anymore), take the non-parameter argument as program name. If multiple are specified, the last one will be picked - don't do that. export TET_RESFILE as <programname>.log by default unless it's already set in the environment. Example use case now: $> pt XGrabButton will generate XGrabButton.log Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net>
2010-05-27xts: Actually wire up the xts shared libraries into the testsDan Nicholson18-54/+18
Commit a2826324 only actually changed the Xproto tests to use the xts shared libraries. This does it for the rest of the tests. On my x86_64 linux box compiling at the default -g -O2, the installed size has gone from 237MB to 72MB. Yay. Signed-off-by: Dan Nicholson <dbn.lists@gmail.com>
2010-05-27tet: Build TET API library as private shared moduleDan Nicholson20-1/+25
Making libapi shared saves quite a bit of space since it's linked into every test program. However, parts of the API are intended to be linked statically into the program, so libapi and libtcm have been split into static and shared (_s) portions. In order to satisy undefined extern variables with shared libraries, tet_pname and tet_thistest have been added to tcc and exported. Hopefully, more code can be moved into the shared _s portion with just small stubs linked statically into the tests. Signed-off-by: Dan Nicholson <dbn.lists@gmail.com>
2010-05-26xts: Remove the logfile make machineryDan Nicholson1-6/+0
Now that the regular test cycle generates both the journal and the summary report, there's really no need for the logfile hack. Signed-off-by: Dan Nicholson <dbn.lists@gmail.com>
2010-05-26xts: Introduce check.sh wrapper script for in-tree testingDan Nicholson1-19/+1
Having all the commands and environment variable overrides for in-tree testing in the Makefile was nice, but a wrapper script means you can do in-tree testing with or without the "make check" machinery. Signed-off-by: Dan Nicholson <dbn.lists@gmail.com>
2010-04-20xts5: Fix "format 'blah' expects type 'blah' but argument has type 'blah'" ↵Aaron Plattner6-24/+25
warnings. Signed-off-by: Aaron Plattner <aplattner@nvidia.com> Acked-by: Peter Hutterer <peter.hutterer@who-t.net>
2010-04-20xts5: Fix "unused variable" warnings.Aaron Plattner236-902/+189
There are still two left: XDrawRectangle/XDrawRectangle.c:4211: warning: unused variable ‘jssav’ XDrawRectangles/XDrawRectangles.c:4292: warning: unused variable ‘jssav’ These are a little harder to fix because the declaration is generated but jssav is only used in some of the tests. Signed-off-by: Aaron Plattner <aplattner@nvidia.com> Acked-by: Peter Hutterer <peter.hutterer@who-t.net>
2010-04-20xts5: Fix "missing braces around initializer" warnings.Aaron Plattner14-511/+510
Signed-off-by: Aaron Plattner <aplattner@nvidia.com> Acked-by: Peter Hutterer <peter.hutterer@who-t.net>
2010-03-24xts5: Make sure libXR5 preceeds the libraries it depends upon in the link orderJon TURNEY3-6/+6
Signed-off-by: Jon TURNEY <jon.turney@dronecode.org.uk> Signed-off-by: Dan Nicholson <dbn.lists@gmail.com>
2010-03-24xts5: Remove an unneeded $(EXEEXT)Jon TURNEY19-19/+19
tprogs lists the test programs without $(EXEEXT), so substituting the $(EXEEXT) with .c to get BUILD_SOURCES fails to work correctly when $(EXEEXT) isn't empty Signed-off-by: Jon TURNEY <jon.turney@dronecode.org.uk> Signed-off-by: Dan Nicholson <dbn.lists@gmail.com>
2010-03-24xts-run: Handle -o parameter as output base directoryDan Nicholson1-4/+5
Allows moving the location of the results directory without having to recreate the dated subdirectory. For example, "xts-run -o ~/xts-results" will give you results in ~/xts-results/$date and subsequent runs will not conflict.
2010-03-24xts5: Tweak the default output directoryDan Nicholson2-2/+2
Collect all the results under one directory instead of cluttering the current directory. Also, add the seconds to the date in the directory name to reduce the chance of conflict.
2010-02-20xts5: Include summary report in resultsDan Nicholson2-3/+19
Give users a quick summary of the test results with the xts-report "no details" mode. This will appear next to the journal as summary.
2010-02-20xts5: Use package version in xts utilitiesDan Nicholson3-2/+15
Use the package version for reporting the version of xts-run and xts-report.
2010-02-19xts5: Fix mc silent rules like AM_V_GENDan Nicholson1-5/+5
I still can't find the issue, but AM_V_mc_ was causing make to stop on a recursive variable. Now it's basically AM_V_GEN with s/GEN/MC/.
2010-02-17xts5: Rename report generator to xts-report and fix some garbageDan Nicholson4-29/+36
xts-report makes a lot more sense than vswrpt since everything else is called xts and you probably couldn't even figure out what vswrpt is. A few fixes along the way: - use DEFAULT_TET_ROOT when $TET_ROOT is not set and use it throughout the code - fix the argv[0] scanner to also accept xts-report - comment out a random form feed (wtf?) and a useless error message - install the docs so xts-report can strip the release data
2010-02-17xts5: Fix comment broken by RCS ID removalDan Nicholson1-0/+1
2010-02-17xts5: Build and install vswrpt report analyzerDan Nicholson5-1/+4
This will be renamed to xts-report, but it has some clever argv[0] matching that prevents it for now. Also fixes a broken comment from an RCS ID removal.
2010-02-17xts5: Set check output directory like xts-runDan Nicholson1-1/+1
2010-02-17xts5: Add some help text to xts-runDan Nicholson1-3/+16
2010-02-17xts5: Use shared library modules to decrease test program sizeDan Nicholson30-97/+109
Linking the xts libraries statically into each test program results in a massive size for the collection of programs. The libraries have been changed to private libtool modules that default to shared, resulting in a much smaller installed size. One consequence of linking to shared objects is that the linker errors on undefined extern variables. The obsolete and unused startup routines in linkstart.c had one of these, so they have been removed from the build. Next is to create the tet library shared.
2010-02-17xts5: Specify default tcc to use in xts-runDan Nicholson1-1/+1
We want to use the tcc built with xts, so default to setting the full path instead of hoping to find it in $PATH.
2010-02-17xts5: Use XTEST extension for testsDan Nicholson1-1/+1
More tests can be run when using XTEST.
2009-10-14xts5: Drop XP pkg-config checkDan Nicholson3-5/+5
The XP check for protocol tests in configure was grabbing the same packages as the general XTS check. Consolidate them.
2009-10-14xts5: Stop gratuitous use of make variables in common.mkDan Nicholson39-196/+315
The meanings of the variables were confusing or obfuscating. Now it's much clearer which libraries are in use.
2009-10-11xts5: Ignore generated xts-config scriptDan Nicholson1-0/+1
2009-10-10xts5: Add more standard headers to kill implicit declaration warningsDan Nicholson11-18/+29
For some reason, there seems to be an aversion to the standard headers. Mostly this is <string.h>, which unfortunately causes a symbol conflict with index on glibc.
2009-10-10xts-run: No need to pass journal fileDan Nicholson1-1/+1
The journal file will default to $results_dir/journal, which is what we want anyway.
2009-10-10tet: Allow for executables outside of TET_ROOT by defaultDan Nicholson1-4/+1
tet is really built around the notion that all the source, programs, data and output will all live in a single tree. However, often times the test programs are built and/or installed in a separate location. This can be worked around with various environment variables or command line options, but it would be better to build this behavior in by default. This introduces TET_EXEC_ROOT, which specifies the base directory for tet test programs. The tests will be found in the test suite subdirectory of TET_EXEC_ROOT. E.g., $TET_EXEC_ROOT/xts. The default here is to have TET_ROOT=$datadir and TET_EXEC_ROOT=$libexecdir.
2009-10-10xts5: Also install tet_codeDan Nicholson1-1/+1
2009-10-10Use a default TET_ROOT when not in the environmentDan Nicholson3-11/+16
Requiring an environment variable to be set in order to get any behavior is an obnoxious trait. Instead, a sane default TET_ROOT is built into the programs that use it. Users such as "make check" that need more control can continue using the environment variable. The scripts are now substituted from make so that we get full variable resolution.
2009-10-10xts5: Install the test fontsDan Nicholson1-1/+2
I'm not sure any of the currently installed tests use the test fonts or not, but tetexec.cfg at least creates a fontpath for them.
2009-10-10xts5: Use xts-run wrapper script for executing testsDan Nicholson3-1/+77
xts-run is a simple wrapper script to generate a tetexec.cfg and run tcc with the appropriate arguments.
2009-10-10xts5: Install xts-config for generating tetexec.cfgDan Nicholson1-0/+1
2009-10-07xts5: Install test data filesDan Nicholson3-0/+9
Many of the tests use data files for verifying pixels in images.
2009-10-07Remove unnecessary fileDan Nicholson1-0/+0
2009-10-07xts5: Use temporary tetexec.cfg in case of failuresDan Nicholson1-1/+1
2009-10-07xts5: Silence tetexec.cfg generationDan Nicholson1-1/+1
2009-10-07Remove DECNET configuration items.Peter Hutterer3-7/+1
Welcome to the wooorld of tomorrow!
2009-10-07xts5: Don't regenerate tetexec.cfg after Makefile updatedDan Nicholson1-4/+2
Before when the display number was set by configure, we wanted to make sure tetexec.cfg was updated with the Makefile. That's not necessary anymore.
2009-10-07Purge XT_DISPLAY config variable, use DISPLAY environment var instead.Peter Hutterer22-83/+69
Let's use DISPLAY like every other sane X application does, not through a variable set at configure time. Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> Signed-off-by: Dan Nicholson <dbn.lists@gmail.com>
generated by cgit v1.2.3 (git 2.39.1) at 2024-12-03 17:48:07 +0000