This is an auto-filled bug because dev-util/dejagnu fails tests. The issue was originally discovered on arm64, but it may be reproducible on other arches as well. If you think that a different summary clarifies the issue better, feel free to change it. Attached build log and emerge --info.
Created attachment 635732 [details] build.log build log and emerge --info
> ERROR: tcl error sourcing ./testsuite/libdejagnu/tunit.exp. > ERROR: The system has no more ptys. Ask your system administrator to create more. Make sure your PTYs work. The simple test is usually to run screen and tmux to see if it creates ttys.
is this what we are looking for? $ ls /dev/ttyS* -la crw-rw---- 1 root dialout 4, 64 Feb 8 12:34 /dev/ttyS0 crw-rw---- 1 root dialout 4, 65 Feb 8 12:34 /dev/ttyS1 crw-rw---- 1 root dialout 4, 66 Feb 8 12:34 /dev/ttyS2 crw-rw---- 1 root dialout 4, 67 Feb 8 12:34 /dev/ttyS3
(In reply to Agostino Sarubbo from comment #3) > is this what we are looking for? > > $ ls /dev/ttyS* -la > crw-rw---- 1 root dialout 4, 64 Feb 8 12:34 /dev/ttyS0 > crw-rw---- 1 root dialout 4, 65 Feb 8 12:34 /dev/ttyS1 > crw-rw---- 1 root dialout 4, 66 Feb 8 12:34 /dev/ttyS2 > crw-rw---- 1 root dialout 4, 67 Feb 8 12:34 /dev/ttyS3 No, these are serial ports. 'expect' library uses pseudo-terminals in /dev/pts. Valid setup usually looks like: $ LANG=C ls -l /dev/ptmx crw-rw-rw- 1 root tty 5, 2 May 3 17:21 /dev/ptmx And as a user (portage user in this case) you can create new ttys: $ tty /dev/pts/14 $ tmux $ tty /dev/pts/15 $ LANG=C ls -l /dev/pts total 0 crw--w---- 1 slyfox tty 136, 0 May 3 17:24 0 crw--w---- 1 slyfox tty 136, 1 May 3 17:24 1 crw--w---- 1 slyfox tty 136, 10 May 2 22:05 10 crw--w---- 1 slyfox tty 136, 11 May 3 15:26 11 crw--w---- 1 slyfox tty 136, 12 May 3 14:45 12 crw--w---- 1 slyfox tty 136, 13 May 3 15:26 13 crw--w---- 1 slyfox tty 136, 14 May 3 17:24 14 crw--w---- 1 slyfox tty 136, 2 May 3 17:24 2 crw--w---- 1 slyfox tty 136, 3 May 3 13:28 3 crw--w---- 1 slyfox tty 136, 4 May 3 13:27 4 crw--w---- 1 slyfox tty 136, 5 May 3 14:32 5 crw--w---- 1 slyfox tty 136, 6 May 2 22:05 6 crw--w---- 1 slyfox tty 136, 7 May 3 17:24 7 crw--w---- 1 slyfox tty 136, 9 May 3 17:24 9 c--------- 1 root root 5, 2 Apr 18 20:31 ptmx
I can create tty as user, however, the /dev/ptmx output looks a bit different: $ LANG=C ls -l /dev/ptmx lrwxrwxrwx 1 root root 8 May 2 11:35 /dev/ptmx -> pts/ptmx $ LANG=C ls -l /dev/pts/ptmx c--------- 1 root root 5, 2 May 4 16:33 /dev/pts/ptmx
(In reply to Agostino Sarubbo from comment #5) > I can create tty as user, however, the /dev/ptmx output looks a bit > different: > > $ LANG=C ls -l /dev/ptmx > lrwxrwxrwx 1 root root 8 May 2 11:35 /dev/ptmx -> pts/ptmx > > $ LANG=C ls -l /dev/pts/ptmx > c--------- 1 root root 5, 2 May 4 16:33 /dev/pts/ptmx Can you run 'tmux' as portage user? # su -s /bin/bash - portage $ tmux $ tty /dev/pts/5 $ LANG=C ls -l /dev/pts/5 crw--w---- 1 portage tty 136, 5 May 4 18:00 /dev/pts/5
This issue comes up because an hardware issue so this is invalid. Do you think it can be useful a portage feature that tests if it is able to create a pty before emerge?