Test reports needed for release 3.0rc1 tarball

GNU Radio
1

This is just a reminder to the group that we have GNU Radio release
3.0rc1 available for testing:

http://gnuradio.org/releases/gnuradio/gnuradio-3.0rc1.tar.gz

There have been no reported problems so far, but it would also be
helpful to get more “works for me” reports as well. You can send a note
to me or to the list.

Once the tarball is unpacked, the only difference in build instructions
between the tarball and building from SVN is that one does not need to
run ‘bootstrap’ first.

The release tarball does not include all the components that are
available via SVN, as some are considered experimental, incomplete, or
destined for a future ‘contrib’ tarball. You can still obtain these via
SVN.

Johnathan C., AE6HO
[email protected]

2

I haven’t tried the tarball, but checking the latest trunk from SVN
worked
fine. I’ve only played with the digital modulation portions so far, but
no
problem with the install or operation of anything I’ve used after a make
distclean.

Tom

3

G’day,

3.0rc1 fails on NetBSD due to missing fusb_ra_wb.h:

gmake[5]: Entering directory `/tmp/gnuradio-3.0rc1/usrp/host/lib’
if /bin/ksh …/…/…/libtool --tag=CXX --mode=compile
g++ -DHAVE_CONFIG_H -I. -I. -I…/…/… -I…/…/…/usrp/host/lib
-I…/…/…/usrp/firmware/include -I/usr/pkg/include -I/usr/pkg/include
-Wall -Woverloaded-virtual -pthread -MT
fusb_ra_wb.lo -MD -MP -MF “.deps/fusb_ra_wb.Tpo” -c -o fusb_ra_wb.lo
fusb_ra_wb.cc;
then mv -f “.deps/fusb_ra_wb.Tpo” “.deps/fusb_ra_wb.Plo”; else
rm -f “.deps/fusb_ra_wb.Tpo”; exit 1; fi

g++ -DHAVE_CONFIG_H -I. -I. -I…/…/… -I…/…/…/usrp/host/lib
-I…/…/…/usrp/firmware/include -I/usr/pkg/include -I/usr/pkg/include
-Wall -Woverloaded-virtual -pthread -MT
fusb_ra_wb.lo -MD -MP -MF .deps/fusb_ra_wb.Tpo -c
fusb_ra_wb.cc -fPIC -DPIC -o .libs/fusb_ra_wb.o
fusb_ra_wb.cc:27:24: error: fusb_ra_wb.h: No such file or directory
fusb_ra_wb.cc:78: error: ‘fusb_devhandle_ra_wb’ has not been declared
fusb_ra_wb.cc:78: error: ISO C++ forbids declaration of
‘fusb_devhandle_ra_wb’
with no type
[…]

The patches that optimizes data bandwidth on USB interface hasn’t made
it into
the trunk :frowning:

cheerio Berndt

4

Discuss-gnuradio mailing list
[email protected]
http://lists.gnu.org/mailman/listinfo/discuss-gnuradio

5

On Tue, Oct 03, 2006 at 12:47:05PM -0400, Tom R. wrote:

I haven’t tried the tarball, but checking the latest trunk from SVN worked
fine. I’ve only played with the digital modulation portions so far, but no
problem with the install or operation of anything I’ve used after a make
distclean.

Tom

Thanks.

However, we’re really looking for feedback on the tarball.
If we don’t hear any, we’ll assume that everything is OK, which it
might not be :wink:

Please let us know now, before we make the official 3.0 release.

Eric

6

Discuss-gnuradio mailing list
[email protected]
http://lists.gnu.org/mailman/listinfo/discuss-gnuradio

7

Greg T. wrote:

The problem is that fusb_ra_wb.h isn’t listed in noinst_HEADERS. I
suppose make distcheck would catch this in svn if run on recent-enough
NetBSD. I’ll run that tomorrow.

This has been fixed in r3697 on the SVN trunk and r3698 in the
release-3.0 branch, which will eventually make it into a new 3.0rc2
tarball.

Johnathan C., AE6HO
[email protected]

8

On Tue, Oct 03, 2006 at 03:59:47PM -0400, Greg T. wrote:

The problem is that fusb_ra_wb.h isn’t listed in noinst_HEADERS. I
suppose make distcheck would catch this in svn if run on recent-enough
NetBSD. I’ll run that tomorrow.

Greg T. <[email protected]>

Could also have been caught using the check-tarball-h-files script
contained in dtools/bin.

Eric

9

Michael D. wrote:

This is for 3.0rc1, trying to compile under OSX 10.4.7 and 10.4.8 on
both PPC and Intel Macs. Everything basically configures (except see
below), makes (“make” and “make -j3”), installs, and distchecks. I will
try actually running (3.0rc2) tomorrow on a known-working OSX PPC (dual
G5). - MLD

Thanks.

checking for machine dependent speedups… generic
…/configure: line 28595: test: =: unary operator expected

Can you send me your ./configure file? I seem to recall this came up in
a different place in the past and we fixed it.

gr-video-sdl requires library sdl, not found.

The configure script will say this for any reason the SDL test fails, so
it should probably say ‘not found or failed compilation test’ or
something like that to be more accurate.

Johnathan C., AE6HO
[email protected]

10

Johnathan C. wrote:

Can you send me your ./configure file? I seem to recall this came up in
a different place in the past and we fixed it.

Um, never mind.

-Johnathan

11

This is for 3.0rc1, trying to compile under OSX 10.4.7 and 10.4.8 on
both PPC and Intel Macs. Everything basically configures (except see
below), makes (“make” and “make -j3”), installs, and distchecks. I
will try actually running (3.0rc2) tomorrow on a known-working OSX
PPC (dual G5). - MLD

In configure:
(1) The “machine dependent speedups” looks like it has a testing bug;
this doesn’t break things, but it does output a warning:

checking for machine dependent speedups… generic
…/configure: line 28595: test: =: unary operator expected
checking for cppunit-config… /opt/local/bin/cppunit-config

(2) Checking for sdl doesn’t work on any of these Macs —when done
via a remote login—; it requires an X11 display, which no matter
how hard I try I can’t get to work reliably even using “ssh -X” etc;
it does work when done by the current “console” user on any X11 xterm.

checking for sdl-config… /opt/local/bin/sdl-config
checking for SDL - version >= 1.2.0… no
*** Could not run SDL test program, checking why…
*** The test program compiled, but did not run. This usually means
*** that the run-time linker is not finding SDL or finding the wrong
*** version of SDL. If it is not finding SDL, you’ll need to set your
*** LD_LIBRARY_PATH environment variable, or edit /etc/ld.so.conf to
point
*** to the installed location Also, make sure you have run ldconfig
if that
*** is required on your system

*** If you have an old version installed, it is best to remove it,
although
*** you may also be able to get things to work by modifying
LD_LIBRARY_PATH
gr-video-sdl requires library sdl, not found.

% sdl-config --version
1.2.11

12

gr-video-sdl requires library sdl, not found.

The configure script will say this for any reason the SDL test
fails, so
it should probably say ‘not found or failed compilation test’ or
something like that to be more accurate.

Not sure if it helps, but here’s the relevant output from config.log.

configure:33616: checking for sdl-config
configure:33634: found /opt/local/bin/sdl-config
configure:33647: result: /opt/local/bin/sdl-config
configure:33656: checking for SDL - version >= 1.2.0
configure:33751: gcc -o conftest -g -O2 -Wall -I/opt/local/include/
SDL -D_GNU_SOURCE=1 -D_THREAD_SAFE -L/opt/local/lib conftest.c -L/
opt/local/lib -lSDLmain -lSDL -Wl,-framework,Cocoa >&5
configure:33754: $? = 0
configure:33760: ./conftest
kCGErrorRangeCheck : Window Server communications from outside of
session allowed for root and console user only
INIT_Processeses(), could not establish the default connection to the
WindowServer…/configure: line 1: 3313 Abort trap ./
conftest$ac_exeext
configure:33763: $? = 134
configure: program exited with status 134
configure: failed program was:
| /* confdefs.h. /
| #define PACKAGE_NAME “”
| #define PACKAGE_TARNAME “”
| #define PACKAGE_VERSION “”
| #define PACKAGE_STRING “”
| #define PACKAGE_BUGREPORT “”
| #define PACKAGE “gnuradio”
| #define VERSION “3.0rc1”
| #define STDC_HEADERS 1
| #define HAVE_SYS_TYPES_H 1
| #define HAVE_SYS_STAT_H 1
| #define HAVE_STDLIB_H 1
| #define HAVE_STRING_H 1
| #define HAVE_MEMORY_H 1
| #define HAVE_STRINGS_H 1
| #define HAVE_INTTYPES_H 1
| #define HAVE_STDINT_H 1
| #define HAVE_UNISTD_H 1
| #define HAVE_DLFCN_H 1
| #define HAVE_PYTHON_H 1
| #define OMNITHREAD_POSIX 1
| #define HAVE_GETTIMEOFDAY 1
| #define HAVE_NANOSLEEP 1
| #define HAVE_SYS_IPC_H 1
| #define HAVE_SYS_SHM_H 1
| #define STDC_HEADERS 1
| #define HAVE_SYS_WAIT_H 1
| #define HAVE_FCNTL_H 1
| #define HAVE_LIMITS_H 1
| #define HAVE_STRINGS_H 1
| #define HAVE_TIME_H 1
| #define HAVE_SYS_IOCTL_H 1
| #define HAVE_SYS_TIME_H 1
| #define HAVE_UNISTD_H 1
| #define HAVE_SYS_MMAN_H 1
| #define HAVE_SYS_SELECT_H 1
| #define HAVE_SYS_TYPES_H 1
| #define HAVE_SYS_RESOURCE_H 1
| #define HAVE_STDINT_H 1
| #define HAVE_SCHED_H 1
| #define TIME_WITH_SYS_TIME 1
| #define WORDS_BIGENDIAN 1
| #define HAVE_ALLOCA_H 1
| #define HAVE_ALLOCA 1
| #define PROTOTYPES 1
| #define __PROTOTYPES 1
| #define HAVE_VPRINTF 1
| #define HAVE_MMAP 1
| #define HAVE_SELECT 1
| #define HAVE_SOCKET 1
| #define HAVE_STRCSPN 1
| #define HAVE_STRERROR 1
| #define HAVE_STRSPN 1
| #define HAVE_GETPAGESIZE 1
| #define HAVE_SYSCONF 1
| #define HAVE_SNPRINTF 1
| #define HAVE_GETTIMEOFDAY 1
| #define HAVE_NANOSLEEP 1
| #define HAVE_MODF 1
| #define HAVE_SQRT 1
| #define HAVE_SINF 1
| #define HAVE_COSF 1
| #define HAVE_TRUNC 1
| #define HAVE_SHM_OPEN 1
| #define HAVE_SYS_TYPES_H 1
| #define HAVE_FCNTL_H 1
| #define HAVE_GETOPT 1
| #define HAVE_USLEEP 1
| #define HAVE_GETTIMEOFDAY 1
| #define HAVE_NANOSLEEP 1
| #define HAVE_RAND 1
| #define HAVE_SRAND 1
| #define HAVE_RANDOM 1
| #define HAVE_SRANDOM 1
| #define HAVE_SLEEP 1
| #define HAVE_SIGACTION 1
| #define HAVE_STRUCT_TIMEZONE 1
| #define HAVE_STRUCT_TIMESPEC 1
| #define HAVE_SSIZE_T 1
| #define HAVE_GETOPT 1
| #define HAVE_USLEEP 1
| #define HAVE_GETTIMEOFDAY 1
| #define HAVE_FCNTL_H 1
| #define HAVE_SYS_TIME_H 1
| #define HAVE_UNISTD_H 1
| #define HAVE_DECL_GETENV 1
| #define HAVE_MKSTEMP 1
| #define WORDS_BIGENDIAN 1
| #define HAVE_GETRUSAGE 1
| #define HAVE_SIGACTION 1
| #define HAVE_SNPRINTF 1
| #define HAVE_USB_H 1
| #define HAVE_AUDIOUNIT_AUDIOUNIT_H 1
| #define HAVE_AUDIOTOOLBOX_AUDIOTOOLBOX_H 1
| / end confdefs.h. /
|
| #include <stdio.h>
| #include <stdlib.h>
| #include <string.h>
| #include “SDL.h”
|
| char
| my_strdup (char str)
| {
| char new_str;
|
| if (str)
| {
| new_str = (char )malloc ((strlen (str) + 1) * sizeof(char));
| strcpy (new_str, str);
| }
| else
| new_str = NULL;
|
| return new_str;
| }
|
| int main (int argc, char argv[])
| {
| int major, minor, micro;
| char tmp_version;
|
| / This hangs on some systems (?)
| system (“touch conf.sdltest”);
| /
| { FILE fp = fopen(“conf.sdltest”, “a”); if ( fp ) fclose(fp); }
|
| / HP/UX 9 (%@#!) writes to sscanf strings /
| tmp_version = my_strdup(“1.2.0”);
| if (sscanf(tmp_version, “%d.%d.%d”, &major, &minor, &micro) != 3) {
| printf("%s, bad version string\n", “1.2.0”);
| exit(1);
| }
|
| if ((1 > major) ||
| ((1 == major) && (2 > minor)) ||
| ((1 == major) && (2 == minor) && (11 >= micro)))
| {
| return 0;
| }
| else
| {
| printf("\n ‘sdl-config --version’ returned %d.%d.%d, but
the minimum version\n", 1, 2, 11);
| printf(" of SDL required is %d.%d.%d. If sdl-config is
correct, then it is\n", major, minor, micro);
| printf("*** best to upgrade to the required version.\n");
| printf("*** If sdl-config was wrong, set the environment
variable SDL_CONFIG\n");
| printf("*** to point to the correct copy of sdl-config, and
remove the file\n");
| printf("*** config.cache before re-running configure\n");
| return 1;
| }
| }
|
|
configure:33787: result: no
configure:33830: gcc -o conftest -g -O2 -Wall -I/opt/local/include/
SDL -D_GNU_SOURCE=1 -D_THREAD_SAFE -L/opt/local/lib conftest.c -L/
opt/local/lib -lSDLmain -lSDL -Wl,-framework,Cocoa >&5
configure:33836: $? = 0
configure:33843: test -z “$ac_c_werror_flag” || test ! -s conftest.err
configure:33846: $? = 0
configure:33853: test -s conftest
configure:33856: $? = 0
configure:33885: result: gr-video-sdl requires library sdl, not found.
configure:33901: result: Not building component gr-video-sdl.

13

This has been fixed in r3697 on the SVN trunk and r3698 in the
release-3.0 branch, which will eventually make it into a new 3.0rc2
tarball.

Thanks. I just ran ‘make distcheck’ on a pretty-recent NetBSD system
with up-to-date svn and ran into only a few problems (the first two of
which aren’t new, but I’m listing them to be helpful to anyone else
trying this):

  1. One needs to give LDFLAGS and CPPFLAGS in the environment; the
    values from original configure aren’t propagated to the configure
    run as part of distcheck.

  2. The test of gr_vmcircbuf_mmap_tmpfile_factory uses vast amount of
    space in /tmp. My /tmp is on mfs and 64 MB, so it fails.

  3. One of the test leaves shm turds, apparently 3 from running make
    check. Surely this must be gr_vmcircbuf_sysv_shm_factory, but I
    saw no easy way to run this test separately.

  4. Running with TMP=/home/tmp, gr_vmcircbuf_mmap_tmpfile seems to be
    ok, but then seems to free things incorrectly.

I can try a RC tarball when there’s a new one.

gmake check-TESTS
gmake[1]: Entering directory
`/home/gdt/ADROIT-public/gnuradio/gnuradio-3.0svn/_build/gnuradio-core/src/tests’
test_all in free(): warning: modified (chunk-) pointer.
.Testing gr_vmcircbuf_createfilemapping_factory…
gr_vmcircbuf_createfilemapping: createfilemapping is not available
… gr_vmcircbuf_createfilemapping_factory: Doesn’t work
Testing gr_vmcircbuf_sysv_shm_factory…
gr_vmcircbuf_sysv_shm: shmget (2): Cannot allocate memory
gr_vmcircbuf_sysv_shm: shmget (1): Invalid argument
gr_vmcircbuf_sysv_shm: shmget (1): Cannot allocate memory
… gr_vmcircbuf_sysv_shm_factory: Doesn’t work
Testing gr_vmcircbuf_mmap_shm_open_factory…
gr_vmcircbuf_mmap_shm_open: mmap or shm_open is not available
… gr_vmcircbuf_mmap_shm_open_factory: Doesn’t work
Testing gr_vmcircbuf_mmap_tmpfile_factory…
… gr_vmcircbuf_mmap_tmpfile_factory: OK
test_all in free(): warning: junk pointer, too low to make sense.
test_all in free(): warning: junk pointer, too low to make sense.
.test_all in free(): warning: junk pointer, too low to make sense.
.test_all in free(): warning: junk pointer, too low to make sense.
.test_all in free(): warning: junk pointer, too low to make sense.
.test_all in free(): warning: junk pointer, too low to make sense.
.test_all in free(): warning: junk pointer, too low to make sense.
.test_all in free(): warning: junk pointer, too low to make sense.
.test_all in free(): warning: junk pointer, too low to make sense.
.test_all in free(): warning: junk pointer, too low to make sense.
.test_all in free(): warning: modified (chunk-) pointer.
test_all in free(): warning: modified (chunk-) pointer.
test_all in free(): warning: modified (chunk-) pointer.
test_all in free(): warning: modified (chunk-) pointer.
test_all in free(): warning: modified (chunk-) pointer.
test_all in free(): warning: modified (chunk-) pointer.
test_all in free(): warning: junk pointer, too low to make sense.
.test_all in free(): warning: modified (chunk-) pointer.
test_all in free(): warning: modified (chunk-) pointer.
test_all in free(): warning: modified (chunk-) pointer.
test_all in free(): warning: modified (chunk-) pointer.
test_all in free(): warning: page is already free.
test_all in free(): warning: junk pointer, too low to make sense.
.test_all in free(): warning: junk pointer, too low to make sense.
.test_all in free(): warning: page is already free.
test_all in free(): warning: page is already free.
[many such lines removed]
test_all in free(): warning: page is already free.
test_all in free(): warning: page is already free.
test_all in free(): warning: page is already free.
test_all in free(): warning: junk pointer, too low to make sense.
.test_all in free(): warning: junk pointer, too low to make sense.
.test_all in free(): warning: junk pointer, too low to make sense.
.test_all in free(): warning: junk pointer, too low to make sense.
Memory fault (core dumped)
FAIL: test_all

1 of 1 tests failed

gmake[1]: *** [check-TESTS] Error 1
gmake[1]: Leaving directory
`/home/gdt/ADROIT-public/gnuradio/gnuradio-3.0svn/_build/gnuradio-core/src/tests’
gmake: *** [check-am] Error 2