MFG:Garrison - Ubuntu 16.04.1 dmesg error "systemd[1]: Failed to start Set console keymap."

Default Comment by Bridge

Default Comment by Bridge

------- Comment From <email address hidden> 2016-09-09 06:54 EDT-------
== Comment: #13 - Application Cdeadmin <> - 2016-09-08 09:05:40 ====== State: MoreInfo by: knodelk on 25 August 2016 11:10:22 ====

and /var/log/kern.log is attached==== State: Open by: knodelk on 23 August 2016 14:00:43 ====

== Comment: #5 - Application Cdeadmin <> - 2016-08-25 12:15:34 ====== State: Assigned by: cde00 on 09 September 2016 05:53:29 ====

Thank you for taking the time to report this bug and helping to make Ubuntu better. It seems that your bug report is not filed about a specific source package though, rather it is just filed against Ubuntu in general. It is important that bug reports be filed about source packages so that people interested in the package can find the bugs about it. You can find some hints about determining what package your bug might be about at https://wiki.ubuntu.com/Bugs/FindRightPackage. You might also ask for help in the #ubuntu-bugs irc channel on Freenode.

To change the source package that this bug is filed about visit https://bugs.launchpad.net/ubuntu/+bug/1621824/+editstatus and add the package name in the text box next to the word Package.

[This is an automated message. I apologize if it reached you inappropriately; please just reply to this message indicating so.]

------- Comment From <email address hidden> 2016-09-09 09:13 EDT-------
== Comment: #1 - Application Cdeadmin <> - 2016-08-23 15:05:14 ====== State: MoreInfo by: knodelk on 25 August 2016 11:10:22 ====

== Comment: #5 - Application Cdeadmin <> - 2016-08-25 12:15:34 ====== State: Assigned by: cde00 on 09 September 2016 08:13:00 ====

Default Comment by Bridge

Default Comment by Bridge

> Here is the requested command output:

> # systemctl status -l console-setup.service

This is not the systemd unit producing that message; the correct unit is keyboard-setup. Please instead provide the output of:

# systemctl status -l keyboard-setup.service

------- Comment From <email address hidden> 2016-09-14 18:54 EDT-------
== Comment: #5 - Application Cdeadmin <> - 2016-08-25 12:15:34 ====== State: Assigned by: cde00 on 14 September 2016 17:54:03 ====

------- Comment From <email address hidden> 2016-09-19 09:50 EDT-------
Here is the output of "systemctl status -l keyboard-setup.service"

root@rcx2d407:~# systemctl status -l keyboard-setup.service
* keyboard-setup.service - Set console keymap
Loaded: loaded (/lib/systemd/system/keyboard-setup.service; disabled; vendor preset: enabled)
Active: failed (Result: exit-code) since Thu 2016-09-15 11:24:18 CDT; 4min 59s ago
Process: 9369 ExecStart=/usr/bin/loadkeys /etc/console-setup/cached.kmap.gz (code=exited, status=1/FAILURE)
Main PID: 9369 (code=exited, status=1/FAILURE)

Sep 15 11:24:18 rcx2d407 systemd[1]: Starting Set console keymap...
Sep 15 11:24:18 rcx2d407 loadkeys[9369]: Couldn't get a file descriptor referring to the console
Sep 15 11:24:18 rcx2d407 systemd[1]: keyboard-setup.service: Main process exited, code=exited, status=1/FAILURE
Sep 15 11:24:18 rcx2d407 systemd[1]: Failed to start Set console keymap.
Sep 15 11:24:18 rcx2d407 systemd[1]: keyboard-setup.service: Unit entered failed state.
Sep 15 11:24:18 rcx2d407 systemd[1]: keyboard-setup.service: Failed with result 'exit-code'.
root@rcx2d407:~#

Thanks. The error message in the journal is:

 Sep 15 11:24:18 rcx2d407 loadkeys[9369]: Couldn't get a file descriptor referring to the console

This comes from loadkeys in the 'kbd' package, which carries an Ubuntu-specific patch to open consoles safely. It's possible this "safe" code is not correctly detecting consoles on this device.

Could you attach the output of 'sudo strace /usr/bin/loadkeys /etc/console-setup/cached.kmap.gz' (the command run by /lib/systemd/system/keyboard-setup.service)?

------- Comment From <email address hidden> 2016-09-20 09:38 EDT-------
As requested, here is the output of 'sudo strace /usr/bin/loadkeys /etc/console-setup/cached.kmap.gz'

root@rcx2d407:/home# sudo strace /usr/bin/loadkeys /etc/console-setup/cached.kmap.gz
execve("/usr/bin/loadkeys", ["/usr/bin/loadkeys", "/etc/console-setup/cached.kmap.g"...], [/* 29 vars */]) = 0
brk(NULL) = 0x1003c2d0000
access("/etc/ld.so.nohwcap", F_OK) = -1 ENOENT (No such file or directory)
access("/etc/ld.so.preload", R_OK) = -1 ENOENT (No such file or directory)
open("/usr/local/cuda/lib64/tls/power8/altivec/dfp/libc.so.6", O_RDONLY|O_CLOEXEC) = -1 ENOENT (No such file or directory)
stat("/usr/local/cuda/lib64/tls/power8/altivec/dfp", 0x3fffdf066700) = -1 ENOENT (No such file or directory)
open("/usr/local/cuda/lib64/tls/power8/altivec/libc.so.6", O_RDONLY|O_CLOEXEC) = -1 ENOENT (No such file or directory)
stat("/usr/local/cuda/lib64/tls/power8/altivec", 0x3fffdf066700) = -1 ENOENT (No such file or directory)
open("/usr/local/cuda/lib64/tls/power8/dfp/libc.so.6", O_RDONLY|O_CLOEXEC) = -1 ENOENT (No such file or directory)
stat("/usr/local/cuda/lib64/tls/power8/dfp", 0x3fffdf066700) = -1 ENOENT (No such file or directory)
open("/usr/local/cuda/lib64/tls/power8/libc.so.6", O_RDONLY|O_CLOEXEC) = -1 ENOENT (No such file or directory)
stat("/usr/local/cuda/lib64/tls/power8", 0x3fffdf066700) = -1 ENOENT (No such file or directory)
open("/usr/local/cuda/lib64/tls/altivec/dfp/libc.so.6", O_RDONLY|O_CLOEXEC) = -1 ENOENT (No such file or directory)
stat("/usr/local/cuda/lib64/tls/altivec/dfp", 0x3fffdf066700) = -1 ENOENT (No such file or directory)
open("/usr/local/cuda/lib64/tls/altivec/libc.so.6", O_RDONLY|O_CLOEXEC) = -1 ENOENT (No such file or directory)
stat("/usr/local/cuda/lib64/tls/altivec", 0x3fffdf066700) = -1 ENOENT (No such file or directory)
open("/usr/local/cuda/lib64/tls/dfp/libc.so.6", O_RDONLY|O_CLOEXEC) = -1 ENOENT (No such file or directory)
stat("/usr/local/cuda/lib64/tls/dfp", 0x3fffdf066700) = -1 ENOENT (No such file or directory)
open("/usr/local/cuda/lib64/tls/libc.so.6", O_RDONLY|O_CLOEXEC) = -1 ENOENT (No such file or directory)
stat("/usr/local/cuda/lib64/tls", 0x3fffdf066700) = -1 ENOENT (No such file or directory)
open("/usr/local/cuda/lib64/power8/altivec/dfp/libc.so.6", O_RDONLY|O_CLOEXEC) = -1 ENOENT (No such file or directory)
stat("/usr/local/cuda/lib64/power8/altivec/dfp", 0x3fffdf066700) = -1 ENOENT (No such file or directory)
open("/usr/local/cuda/lib64/power8/altivec/libc.so.6", O_RDONLY|O_CLOEXEC) = -1 ENOENT (No such file or directory)
stat("/usr/local/cuda/lib64/power8/altivec", 0x3fffdf066700) = -1 ENOENT (No such file or directory)
open("/usr/local/cuda/lib64/power8/dfp/libc.so.6", O_RDONLY|O_CLOEXEC) = -1 ENOENT (No such file or directory)
stat("/usr/local/cuda/lib64/power8/dfp", 0x3fffdf066700) = -1 ENOENT (No such file or directory)
open("/usr/local/cuda/lib64/power8/libc.so.6", O_RDONLY|O_CLOEXEC) = -1 ENOENT (No such file or directory)
stat("/usr/local/cuda/lib64/power8", 0x3fffdf066700) = -1 ENOENT (No such file or directory)
open("/usr/local/cuda/lib64/altivec/dfp/libc.so.6", O_RDONLY|O_C...

Thanks, this strace shows that src/getfd.c:getfd_keymap() is unable to find any VTs not in raw mode:

int
getfd_keymap(void) {
        int fd, bestfd = -1;
        char ttyname[sizeof("/dev/ttyNN")] = "/dev/tty";
        int i;

#define CHECK_FD_KEYMAP do { \
        if (is_a_console(fd)) { \
                long mode = kbmode(fd); \
                if (mode == K_UNICODE) { \
                        if (bestfd != -1) \
                                close(bestfd); \
                        return fd; \
                } else if (mode == K_XLATE && bestfd == -1) \
                        bestfd = fd; \
                else \
                        close(fd); \
        } else \
                close(fd); \
} while (0)

[...]

        for (i = 1; i <= 12; ++i) {
                snprintf(ttyname + sizeof("/dev/tty") - 1, 3, "%d", i);
                if ((fd = open_a_console(ttyname)) >= 0)
                        CHECK_FD_KEYMAP;
        }

        fprintf(stderr,
                _("Couldn't get a file descriptor referring to the console\n"));

        /* total failure */
        exit(1);
}

is_a_console() checks KDGKBTYPE, kbmode() checks KDGKBMODE.

Could you please compile and run the attached program as root on the affected system, and post the results?

In the course of investigating, I've discovered locally that ioctl(fd, KDGKBMODE), despite taking an unsigned long * as an argument, doesn't fully clear the address here on xenial x86_64. So if this program returns sensible results on your system, we may be looking at a memory initialization bug.

And the ioctl_list(2) manpage claims that KDGKBMODE should take an int*, not a long*. That would definitely explain this not working on ppc64el.

------- Comment From <email address hidden> 2016-09-20 13:23 EDT-------
cde00 (<email address hidden>) added native attachment /tmp/AIXOS06182883/check-kbd-mode.c on 2016-09-20 12:13:43

And I've just confirmed that with the exception of this Ubuntu patch, all calls to ioctl(..., KDGKBMODE) in the kbd source are using int*, not long*. So this is a bug in that patch, which surprisingly didn't bite us on x86_64 previously but definitely means keymap setting will be heavily dependent on memory initialization.

Results from the test program might still be interesting in case there's another bug here, but I've at least identified /a/ bug that needs fixing.

------- Comment From <email address hidden> 2016-09-22 09:56 EDT-------
output of check-kbd-mode as requested:

root@rcx2d407:/home# ls -l
total 20
-rwxr-xr-x 1 root root 70496 Sep 22 08:53 check-kbd-mode
-rwxrwxrwx 1 root root 1508 Sep 22 08:52 check-kbd-mode.c
root@rcx2d407:/home# check-kbd-mode
KBMODE of /dev/tty1: 3
KBMODE of /dev/tty2: 1
KBMODE of /dev/tty3: 1
KBMODE of /dev/tty4: 1
KBMODE of /dev/tty5: 1
KBMODE of /dev/tty6: 1
KBMODE of /dev/tty7: 3
KBMODE of /dev/tty8: 3
KBMODE of /dev/tty9: 3
KBMODE of /dev/tty10: 3
KBMODE of /dev/tty11: 3
KBMODE of /dev/tty12: 3
root@rcx2d407:/home#

On Thu, Sep 22, 2016 at 02:03:47PM -0000, bugproxy wrote:
> ------- Comment From <email address hidden> 2016-09-22 09:56 EDT-------
> output of check-kbd-mode as requested:

Thanks, that confirms that the bug is the one I found here. You should find
that this is fixed for you in the pending SRU.

Hello bugproxy, or anyone else affected,

Accepted kbd into xenial-proposed. The package will build now and be available at https://launchpad.net/ubuntu/+source/kbd/1.15.5-1ubuntu5 in a few hours, and then in the -proposed repository.

Please help us by testing this new package. See https://wiki.ubuntu.com/Testing/EnableProposed for documentation how to enable and use -proposed. Your feedback will aid us getting this update out to other Ubuntu users.

If this package fixes the bug for you, please add a comment to this bug, mentioning the version of the package you tested, and change the tag from verification-needed to verification-done. If it does not fix the bug for you, please add a comment stating that, and change the tag to verification-failed. In either case, details of your testing will help us make a better decision.

Further information regarding the verification process can be found at https://wiki.ubuntu.com/QATeam/PerformingSRUVerification . Thank you in advance!

This bug was fixed in the package kbd - 2.0.3-2ubuntu2

---------------
kbd (2.0.3-2ubuntu2) yakkety; urgency=medium

  * debian/patches/loadkeys_find_usable_console.diff: fix wrong type argument
    to ioctl(..., KDGKBMODE), which causes consoles to not be correctly
    detected on 64-bit archs. LP: #1621824.

 -- Steve Langasek <email address hidden> Tue, 20 Sep 2016 10:53:30 -0700

This bug was fixed in the package kbd - 1.15.5-1ubuntu5

---------------
kbd (1.15.5-1ubuntu5) xenial; urgency=medium

  * debian/patches/loadkeys_find_usable_console.diff: fix wrong type argument
    to ioctl(..., KDGKBMODE), which causes consoles to not be correctly
    detected on 64-bit archs. LP: #1621824.

 -- Steve Langasek <email address hidden> Tue, 20 Sep 2016 10:55:50 -0700

The verification of the Stable Release Update for kbd has completed successfully and the package has now been released to -updates. Subsequently, the Ubuntu Stable Release Updates Team is being unsubscribed and will not receive messages about this bug report. In the event that you encounter a regression using the package from -updates please report a new bug using ubuntu-bug and tag the bug report regression-update so we can easily find any regressions.

Hi,

I've same error on my arm64 board with xenial.
Could you also apply this for arm64 architecture?

------- Comment From <email address hidden> 2017-02-01 04:05 EDT-------

@parkch98

All bug fixes are compiled and published for all architectures simultaniously. This bug report is closed, please open a new bug report for your issue. It's best to have new bugs, rather than reopening old issues. Despite similar symptoms it could be an entirely different cause.