My plan once all the BSPs are building again with the tester is to see which are disabled the most across the set of BSPs using xxx-testsuite.tcfg files. That should be be the set of tests which are candidates for review on how to shrink them.
Chris and I chatted about this and my hope is that every psx or sp test (including tm's) can be reworked, split, resources lowered, etc so it can work even on low memory targets. File system tests may be candidates for a way to specify the size of the RAM disk similar to how many tm tests have OPERATION_COUNT. Some tests like fileio will never shrink enough. Reviewing the .tcfg files for common tests disabled for memory use is the first step. --joel On Tue, Mar 12, 2019 at 9:43 AM Gedare Bloom <ged...@rtems.org> wrote: > Is it feasible to triage the failures for out-of-memory? > > On Sat, Mar 9, 2019 at 12:22 PM Joel Sherrill <j...@rtems.org> wrote: > >> >> >> On Sat, Mar 9, 2019, 11:01 AM Chris Johns <chr...@rtems.org> wrote: >> >>> On 10/3/19 2:18 am, j...@rtems.org wrote: >>> > Testing time : 0:03:48.073542 >>> > Average test time: 0:00:00.384610 >>> > >>> > Host >>> > ==== >>> > Linux-3.10.0-862.11.6.el7.x86_64-x86_64-with-centos-7.5.1804-Core >>> (Linux rtbf64c.rtems.com 3.10.0-862.11.6.el7.x86_64 #1 SMP Tue Aug 14 >>> 21:49:04 UTC 2018 x86_64 x86_64) >>> > >>> > Configuration >>> > ============= >>> > Version: 5.0.0.08ef50ff6a3b712cef5bcdeb78c531bc6a3e0f9e >>> > Build : RTEMS_NETWORKING,RTEMS_POSIX_API >>> > Tools : 7.4.0 20181206 (RTEMS 5, RSB >>> 38241392a4f96dabf2d1aba51a43dcb623db4dfb, Newlib 1d35a003f) >>> > >>> > Summary >>> > ======= >>> > >>> > Passed: 583 >>> > Failed: 1 >>> > User Input: 5 >>> > Expected Fail: 0 >>> > Indeterminate: 0 >>> > Benchmark: 3 >>> > Timeout: 1 >>> > Invalid: 0 >>> > Wrong Version: 0 >>> > Wrong Build: 0 >>> > Wrong Tools: 0 >>> > ------------------ >>> > Total: 593 >>> > >>> > Failures: >>> > dl09.exe >>> >>> This test needs a lot of memory to run. It loads modules 64K apart to >>> test the >>> trampolines. I think there is an option to increase the memory of this >>> simulator >>> but I am not sure. >>> >>> I am wondering if we need a file called 'disable >>> -large-memory-tests.tcfg' with >>> dl09.exe excluded. >>> >> >> Some of the tests are known to be deliberate memory pigs. This would be >> good. Commonly excluded tests would be obvious candidates. >> >> I need to investigate the impact of enabling debug and profiling on the >> failures. All built for me without those enabled. >> >>> >>> Chris >>> _______________________________________________ >>> devel mailing list >>> devel@rtems.org >>> http://lists.rtems.org/mailman/listinfo/devel >>> >> _______________________________________________ >> devel mailing list >> devel@rtems.org >> http://lists.rtems.org/mailman/listinfo/devel > >
_______________________________________________ devel mailing list devel@rtems.org http://lists.rtems.org/mailman/listinfo/devel
