Hi Lennart, sorry, this took a bit longer than expected, but I took the time to upgrade to systemd 221, so the results should be a bit closer to the current state than before (which was still using systemd 219).
Inside container (broken shell/fish): speed 38400 baud; rows 60; columns 184; line = 0; intr = ^C; quit = ^\; erase = ^?; kill = ^U; eof = ^D; eol = <undef>; eol2 = <undef>; swtch = <undef>; start = ^Q; stop = ^S; susp = ^Z; rprnt = ^R; werase = ^W; lnext = ^V; flush = ^O; min = 1; time = 0; -parenb -parodd -cmspar cs8 hupcl -cstopb cread -clocal -crtscts -ignbrk -brkint -ignpar -parmrk -inpck -istrip -inlcr -igncr -icrnl -ixon -ixoff -iuclc -ixany -imaxbel iutf8 opost -olcuc -ocrnl onlcr -onocr -onlret -ofill -ofdel nl0 cr0 tab0 bs0 vt0 ff0 isig icanon -iexten echo echoe echok -echonl -noflsh -xcase -tostop -echoprt -echoctl echoke Inside container (working shell/bash): speed 38400 baud; rows 60; columns 184; line = 0; intr = ^C; quit = ^\; erase = ^?; kill = ^U; eof = ^D; eol = <undef>; eol2 = <undef>; swtch = <undef>; start = ^Q; stop = ^S; susp = ^Z; rprnt = ^R; werase = ^W; lnext = ^V; flush = ^O; min = 1; time = 0; -parenb -parodd -cmspar cs8 hupcl -cstopb cread -clocal -crtscts -ignbrk -brkint -ignpar -parmrk -inpck -istrip -inlcr -igncr -icrnl -ixon -ixoff -iuclc -ixany -imaxbel iutf8 opost -olcuc -ocrnl onlcr -onocr -onlret -ofill -ofdel nl0 cr0 tab0 bs0 vt0 ff0 isig icanon -iexten echo echoe echok -echonl -noflsh -xcase -tostop -echoprt -echoctl echoke Outside container: speed 38400 baud; rows 60; columns 184; line = 0; intr = ^C; quit = ^\; erase = ^?; kill = ^U; eof = ^D; eol = <undef>; eol2 = <undef>; swtch = <undef>; start = ^Q; stop = ^S; susp = ^Z; rprnt = ^R; werase = ^W; lnext = ^V; flush = ^O; min = 1; time = 0; -parenb -parodd -cmspar cs8 -hupcl -cstopb cread -clocal -crtscts -ignbrk -brkint -ignpar -parmrk -inpck -istrip -inlcr -igncr -icrnl -ixon -ixoff -iuclc -ixany -imaxbel iutf8 -opost -olcuc -ocrnl onlcr -onocr -onlret -ofill -ofdel nl0 cr0 tab0 bs0 vt0 ff0 -isig -icanon -iexten -echo echoe echok -echonl -noflsh -xcase -tostop -echoprt echoctl echoke Diff inside/outside: opost/-opost isig/-isig icanon/-icanon echo/-echo -echoctl/echoctl I am not 100% sure I did gather this data correctly though: Systemd had already brought up the machine during boot, so I just ran machinectl login and then did the stty commands there. I also did not log out to switch between fish and bash, I just started bash from fish, so I am not surprised that the results are -- to my untrained eye -- identical for these two cases. Running ssty -a -F /dev/pts/2 in another konsole window did produce some differences. It does not seem to effect the result in any way whether or not I run this in konsole or xterm. I got this on a laptop, so if you want to play with a machine that shows this behavior I can demonstrate it if that help:-) Best Regards, Tobias On Fri, Jun 19, 2015 at 10:15 AM, Tobias Hunger <tobias.hun...@gmail.com> wrote: > Thanks for the reply! > > I'll try to collect all requested info tonight or over the weekend. > > Best Regards, > Tobias > > On Thu, Jun 18, 2015 at 9:24 PM, Lennart Poettering > <lenn...@poettering.net> wrote: >> On Tue, 26.05.15 21:40, Tobias Hunger (tobias.hun...@gmail.com) wrote: >> >>> This is stty -a from outside the container: >>> >>> speed 38400 baud; rows 46; columns 114; line = 0; >>> intr = ^C; quit = ^\; erase = ^?; kill = ^U; eof = ^D; eol = M-^?; >>> eol2 = M-^?; swtch = <undef>; start = ^Q; >>> stop = ^S; susp = ^Z; rprnt = ^R; werase = ^W; lnext = ^V; flush = ^O; >>> min = 1; time = 0; >>> -parenb -parodd -cmspar cs8 -hupcl -cstopb cread -clocal -crtscts >>> -ignbrk -brkint -ignpar -parmrk -inpck -istrip -inlcr -igncr icrnl >>> -ixon -ixoff -iuclc ixany imaxbel -iutf8 >>> opost -olcuc -ocrnl onlcr -onocr -onlret -ofill -ofdel nl0 cr0 tab0 bs0 vt0 >>> ff0 >>> isig icanon iexten echo echoe echok -echonl -noflsh -xcase -tostop >>> -echoprt echoctl echoke >>> >>> This is stty -a inside the nspawn-container: >>> >>> speed 38400 baud; rows 46; columns 114; line = 0; >>> intr = ^C; quit = ^\; erase = ^?; kill = ^U; eof = ^D; eol = <undef>; >>> eol2 = <undef>; swtch = <undef>; start = ^Q; >>> stop = ^S; susp = ^Z; rprnt = ^R; werase = ^W; lnext = ^V; flush = ^O; >>> min = 1; time = 0; >>> -parenb -parodd -cmspar cs8 hupcl -cstopb cread -clocal -crtscts >>> -ignbrk -brkint -ignpar -parmrk -inpck -istrip -inlcr -igncr -icrnl >>> -ixon -ixoff -iuclc -ixany -imaxbel iutf8 >>> opost -olcuc -ocrnl onlcr -onocr -onlret -ofill -ofdel nl0 cr0 tab0 bs0 vt0 >>> ff0 >>> isig icanon -iexten echo echoe echok -echonl -noflsh -xcase -tostop >>> -echoprt -echoctl echoke >>> >>> The difference is: >>> eol, eol2, -icrnl -ixany -imaxbel iutf8 -iexten -echoctl >> >> Sorry for the late reply... >> >> Hmm, I think the most interesting info would actually be to see stty >> -a from a working instance, and from a non-working >> instance. I.e. start the container, log into it, type "stty -a" command when >> everything works, and when it doesn't, and let me know the diff of it. >> >> Also, right after doing the "stty -a" in the container, please run the >> same commands on the host, in a seperate xterm, but connect to the >> host side container tty using "stty -a -F /dev/pts/xyz", where >> /dev/pts/xyz is the pts that nspawn itself is running on. >> >> Or to explain it in more steps: >> >> a) open an xterm of some form >> >> b) type "tty" into it, and remember the pty name it responds. It >> should be something like "/dev/pts/xyz". >> >> c) now run systemd-nspawn inside the xterm, and login there, then type >> "stty -a" in it, and save the output that command generated >> somewhere. >> >> d) now, leave everything as it is now, open a second xterm. In it run >> "stty -a -F /dev/pts/xyz", replacing "/dev/pts/xyz" with the pty >> name from step b) and save the output somwhere. >> >> Then, close both xterms. Do these steps once for a container where >> things work, and once for a container where things are borked. Then >> let me know the diffs between the working and non-working outputs from >> both runs of c), as we as the diffs between the working and >> non-working outputs from both runs of d). >> >> Make sure you take the stty snapshots at the exact same states each >> time, because shells and so on tend to toggle some bits of it >> depending on whether they are in the fg or not... >> >> Also, it would be good, to check if different xterm implementations >> (gnome, kde, original xterm) behave differently. >> >> Lennart >> >> -- >> Lennart Poettering, Red Hat _______________________________________________ systemd-devel mailing list systemd-devel@lists.freedesktop.org http://lists.freedesktop.org/mailman/listinfo/systemd-devel