udev: fix deadlock on xenstore access during dom0 boot

During early dom0 boot, before xenstored is started any access to it
through /proc/xen/xenbus (or /dev/xen/xenbus) will wait until xenstored
is started. If that happens in script ordered to start before xenstored,
we have a deadlock. Prevent this by using `xenstore-ls -s` in dom0,
which will fail immediately when xenstored isn't running yet. This fail
isn't a problem, because when xenstored isn't running yet, surely device
isn't attached anywhere.

QubesOS/qubes-issues#1081

(cherry picked from commit cdbcb2eb55)
This commit is contained in:
Marek Marczykowski-Górecki 2016-05-17 22:41:51 +02:00
parent 95eeeef86c
commit a465359a5d
No known key found for this signature in database
GPG Key ID: 063938BA42CFA724

View File

@ -2,6 +2,7 @@
shopt -s nullglob
export LC_CTYPE=en_US.UTF-8
NAME=${DEVNAME#/dev/}
DESC="`echo "${ID_MODEL} (${ID_FS_LABEL})" | iconv -f utf8 -t ascii//TRANSLIT`"
@ -36,9 +37,18 @@ is_used() {
return 1
}
# communicate with xenstored through socket in dom0
# trying to access xenstore before xenstored is started, hang forever (in
# non-killable state), so better fail ('-s' in VM if /proc/xen isn't mounted
# yet) than hang dom0 boot
if [ ! -r /proc/xen/capabilities ] || grep -q control_d /proc/xen/capabilities; then
XENSTORE_LS="xenstore-ls -s"
else
XENSTORE_LS="xenstore-ls"
fi
is_attached() {
dev_hex=$(stat -c %t:%T /dev/$(basename $1))
xenstore-ls backend/vbd | grep -q "physical-device = \"$dev_hex\""
$XENSTORE_LS backend/vbd | grep -q "physical-device = \"$dev_hex\""
}
# update info about parent devices, if any: