On 8 June 2015 at 10:15, Sami Kerola <kerolasa@xxxxxx> wrote: > On 7 June 2015 at 21:25, Sami Kerola <kerolasa@xxxxxx> wrote: >>> - the "script: replay" test hangs :-( >> >> I cannot reproduce the problem, and hence I don't know what might be >> going on. Perhaps the test should not be merged, or maybe it should >> and some debugging ought to be added to see what exactly hangs and >> why. > > My work desktop hung, and it is somehow caused by the 'sleep 0.1' in: And indeed my home gear hung when sleeping a bit longer. > https://github.com/kerolasa/lelux-utiliteetit/commit/c24d5e3903b0872d6ee956c45c0be760fa308a18#diff-2b763ca5b844d35fb0ade37e5dc2f4a1R28 > > If I remove the sleep all works. When I put some other 'not very quick > to run' command in place of the sleep hung again. With 'sleep 0.01' > the execution is successful. I am starting to feel this might be a bug > in script(1) --command option rather than in the test. I will continue > investigation later this evening. Meanwhile all proposals what could > be going on are welcome. Correct guess earlier the day. The --command option had logic issue, that is now fixed and available in my script3 branch. Fix was nothing more than a removal of negation operator. @@ -254,7 +254,7 @@ static void handle_io(struct script_control *ctl, int fd, int i) /* without sync write_output() will write both input & * shell output that looks like double echoing */ fdatasync(ctl->master); - if (!ctl->isterm && !feof(stdin)) { + if (!ctl->isterm && feof(stdin)) { char c = DEF_EOF; write_all(ctl->master, &c, sizeof(char)); } -- Sami Kerola http://www.iki.fi/kerolasa/ -- To unsubscribe from this list: send the line "unsubscribe util-linux" in the body of a message to majordomo@xxxxxxxxxxxxxxx More majordomo info at http://vger.kernel.org/majordomo-info.html
