GNOME Bugzilla – Bug 649669
build failure on Debian sid for mips
Last modified: 2017-06-21 08:25:49 UTC
Hi, beast 0.7.4 failed to build on Debian unstable for mips architecture. An excerpt of the buildlog follows: Testing fft code for size 65536: [****-****-****-****-****-****-] PASS: testfft TEST: lt-testcxx (lt-testcxx:23659): GLib-GObject-CRITICAL **: g_closure_ref: assertion `closure != NULL' failed (lt-testcxx:23659): GLib-GObject-CRITICAL **: g_closure_sink: assertion `closure != NULL' failed (lt-testcxx:23659): BSE-WARNING **: bsecxxbase.cc:124: ignoring invalid signal connection ("|P" != "|Piriiiii") (lt-testcxx:23659): GLib-GObject-CRITICAL **: g_closure_unref: assertion `closure != NULL' failed (lt-testcxx:23659): GLib-GObject-CRITICAL **: g_closure_ref: assertion `closure != NULL' failed (lt-testcxx:23659): GLib-GObject-CRITICAL **: g_closure_sink: assertion `closure != NULL' failed (lt-testcxx:23659): BSE-WARNING **: bsecxxbase.cc:124: ignoring invalid signal connection ("|P" != "r|is") (lt-testcxx:23659): GLib-GObject-CRITICAL **: g_closure_unref: assertion `closure != NULL' failed PASS: testcxx TEST: lt-subnormals Checking if your processor is in 'denormals are zero' (DAZ) mode... no. testing algorithm zap for correctness: [** BSE-TEST:ERROR:subnormals.cc:55:void test_correct_subnormal_elimination(const char*) [with float (* Func)(float) = test2f]: assertion failed: (BSE_FLOAT_IS_SUBNORMAL (value)) /bin/bash: line 5: 23681 Aborted ${dir}$tst FAIL: subnormals TEST: lt-loophandle lt-loophandle[23705]:BSE-TEST-Information:1: check_loop<0,4095> lt-loophandle[23705]:BSE-TEST-Information:1: check_loop<17,379> lt-loophandle[23705]:BSE-TEST-Information:1: check_loop<59,1930> PASS: loophandle TEST: lt-blocktests Running Default Block Ops: [-] BlockFill: [----] BlockCopy: [-----] BlockAdd: [----] BlockSub: [----] BlockMul: [----] BlockScale: [----] BlockRange: [----------------------------------------] BlockSquareSum: [*--*--*--*--*--*--*--*--*--*--*--*--*--*--*--*--*--*--*--*--] TEST: lt-blocktests PASS: blocktests TEST: lt-resamplehandle Resampler C API (FPU): [*-] Resampler Delay Compensation (FPU): [*-*-*-*-*-*-*-*-*-*-] Resampler State Length Info (FPU): [------] ResampleHandle-Up08S (FPU): [---*-*-*-] ResampleHandle-Dn08S (FPU): [---*-*-*-] ResampleHandle-Up12S (FPU): [---*-*-*-] ResampleHandle-Dn12S (FPU): [---*-*-*-] ResampleHandle-Up16S (FPU): [---*-*-*-] ResampleHandle-Dn16S (FPU): [---*-*-*-] ResampleHandle-Up20S (FPU): [---*-*-*-] ResampleHandle-Dn20S (FPU): [---*-*-*-] ResampleHandle-Up24S (FPU): [---*-*-*-] ResampleHandle-Dn24S (FPU): [---*-*-*-] TEST: lt-resamplehandle PASS: resamplehandle TEST: lt-firhandle Highpass Handle (sweep): [----------] Highpass Handle (multichannel): [-*-*--*-*--*-*--*-*--*-*--*-*--*-*--*-*--*-*--*-*-] Highpass Handle (seek): [---------------------------------] Lowpass Handle (sweep): [----------] Lowpass Handle (multichannel): [-*-*--*-*--*-*--*-*--*-*--*-*--*-*--*-*--*-*--*-*-] Lowpass Handle (seek): [---------------------------------] PASS: firhandle make[5]: =================== *** [check-TESTS] Error 1 1 of 9 tests failed =================== make[5]: Leaving directory `/build/buildd2-beast_0.7.4-1-mips-uXlHyX/beast-0.7.4/bse/tests' make[4]: *** [check-am] Error 2 make[4]: Leaving directory `/build/buildd2-beast_0.7.4-1-mips-uXlHyX/beast-0.7.4/bse/tests' make[3]: *** [check-recursive] Error 1 make[3]: Leaving directory `/build/buildd2-beast_0.7.4-1-mips-uXlHyX/beast-0.7.4/bse/tests' make[2]: *** [check-recursive] Error 1 make[2]: Leaving directory `/build/buildd2-beast_0.7.4-1-mips-uXlHyX/beast-0.7.4/bse' make[1]: *** [check-recursive] Error 1 make[1]: Leaving directory `/build/buildd2-beast_0.7.4-1-mips-uXlHyX/beast-0.7.4' dh_auto_test: make -j10 check returned exit code 2 The full buildlog is available here: <https://buildd.debian.org/status/fetch.php?pkg=beast&arch=mips&ver=0.7.4-1&stamp=1304784768>
Thanks for the report. (In reply to comment #0) > beast 0.7.4 failed to build on Debian unstable for mips architecture. > > An excerpt of the buildlog follows: > TEST: lt-subnormals > Checking if your processor is in 'denormals are zero' (DAZ) mode... no. > testing algorithm zap for correctness: [** > BSE-TEST:ERROR:subnormals.cc:55:void test_correct_subnormal_elimination(const > char*) [with float (* Func)(float) = test2f]: assertion failed: > (BSE_FLOAT_IS_SUBNORMAL (value)) > /bin/bash: line 5: 23681 Aborted ${dir}$tst > FAIL: subnormals Stefan, can you please look into adapting your denormals test for MIPs? It's aparently broken in its current form. In the newer GCC docs, it seems to be called "flush to zero": http://gcc.gnu.org/onlinedocs/gcc-3.4.0/gcc/MIPS-Options.html
Closing because Beast isn't regularly built on MIPS anymore.