http://issues.apache.org/SpamAssassin/show_bug.cgi?id=5574
------- Additional Comments From [EMAIL PROTECTED] 2007-08-01 02:05 ------- (In reply to comment #37) > Is t/root_spamd_u_dcc.t supposed to be a test case for this bug? It isn't in > any > of the attachments here. Yes -- I didn't include the test cases in the backported code; I didn't think it would be a requirement to fix the bug and close the test. simply copying the 2 test scripts from trunk's "t" directory will work fine. > On my high-latency satellite connection I fail root_spamd_u_dcc.t unless the > timeout in the last loop of tests is increased from 10 seconds. Setting it to > 30 > seconds works fine for me. The test only takes as long as is necessary when > there is not a timeout, so increasing the limit will not slow down make test > on > systems that don't have a problem. so, adding "dcc_timeout 30" ? ok, I'll add that to the test case. Note that it's irrelevant if DCC fires or not. The bug happens as soon as the DCC helper process is forked -- before the dccproc process is exec()'d. so if the code gets to exec dccproc, and collects *any* result from that process, even a timeout, the bug is fixed. > Running under FreeBSD 6.2 with perl 5.8.8, using svn trunk and branches/3.2 > all > tests passed both with and without the patches. So the patches don't break > anything, but I did not reproduce the problem that they are supposed to fix. I > did have DCC installed and DCC tests enabled, but not Pyzor. > > To reproduce the problem is DCC supposed to be installed to run under some > user > other than root? I just did a default install, which runs DCC as root. You need to have dcc installed (as root is fine). Pyzor is not required. You also need to edit or create t/config and change "run_dcc_tests" to "y". > If everyone who has seen this problem can say that the patch fixes it for them > I'll vote +1 for it, with t/root_spamd_u_dcc.t added in and with it's timeout > bumped up from 10 seconds to 30. ok, I'll redo the patches with the test scripts included, and increased timeout. ------- You are receiving this mail because: ------- You are the assignee for the bug, or are watching the assignee.
