Martin Pitt [2014-11-12 10:10 +0100]: > So it seems this is a problem on armhf/arm64 with udevadm in the > first iteration?
My initial hunch was that cmd->debug was nonzero as it doesn't get explicitly
initialized anywhere. But that isn't the case, and the static structs get
zeroed properly on ARM as well.
Breakpoint 1, run_command (udev=0x6d008, cmd=0x66ab8 <udevadm_info>, argc=3,
argv=0xfffef718) at ../src/udev/udevadm.c:80
80 if (cmd->debug)
(gdb) p *cmd
$2 = {name = 0x471cc "info", cmd = 0x14441 <uinfo>, help = 0x471d4 "query
sysfs or the udev database", debug = 0}
82 log_debug("calling: %s", cmd->name);
(gdb) n
calling: info
But then I realized that log_debug isn't the *udev* logging function,
but the systemd one. Indeed asachi's /proc/cmdline contains "debug"
and thus enables the systemd debug logging. I figure this is the same
as for the arm64 buildd.
I sent a fix to the upstream ML now:
http://lists.freedesktop.org/archives/systemd-devel/2014-November/025165.html
Martin
--
Martin Pitt | http://www.piware.de
Ubuntu Developer (www.ubuntu.com) | Debian Developer (www.debian.org)
signature.asc
Description: Digital signature
