Created attachment 153653 [details] patch Currently 815 cases are reported to fail on FreeBSD 10.1 amd64, compared to 616 on Arch linux. But it still makes sense to have the target to run it in Makefile. Relevant bug report https://sourceware.org/bugzilla/show_bug.cgi?id=18067 , reporting to gdb many linux failures.
Created attachment 153654 [details] check log for FreeBSD 10.1 amd64
Created attachment 153655 [details] check log for Arch linux for comparison
Hi Yuri, thanks for your help. I' like to integrate your patch (with a fix about a missing dependency) for the next gdb 7.9
Thanks! Just please don't add dejagnu as a dependency of gdb, because it is only for check. Notice will suffice.
This will be solved with Bug 200610
@Yuri, This change needs to use TEST_DEPENDS (for dejagnu), do-test (from the new test framework in the tree) and TEST_TARGET (which i believe defaults to check) post-build comment should be also removed since the existence of a do-test target means it can be run at any time and implicitly since it exists. @Luca, setting maintainer-feedback to - means 'feedback not received'. If you want to reject a patch for any reason, use maintainer-approval and set it to - You can mouseover any flag name for tooltip help descriptions on what each value means and can be used for.
(In reply to Kubilay Kocak from comment #6) @koobs, sorry for the misuse of maintainer-feedback flag. I would set maintainer-approval to - to the submitted patch, but it doesn't work: I open the patch, I change the flag and I push the "Submit" button; I did it twice and nothing happened. Maybe it's a permission/credential stuff. A second problem with this bug entry is that it's kinda duplicated: Bug 200610 has a solution to implement tests. I mentioned it in comment #4, but I guess I should be more explicit and use the word "duplicated", because I've not enough right to set relationship status between bug entries.
*** This bug has been marked as a duplicate of bug 200610 ***