polystrap current status - qemu user mode emulation and tzdata
Hi,
some updates to the qemu user mode emulation status and tzdata wrt
polystrap usage.
1. Remove copying of qemu-user-static
-------------------------------------
There was some discussion about how to best choose the qemu-user-static
binary to be copied into the created chroot environment here:
http://lists.debian.org/debian-embedded/2011/06/msg00093.html
but the whole copying is unnecessary with fakeroot/fakechroot as the
qemu-user binary from the host system will be used to execute foreign
binaries. The qemu-user-static binary inside the chroot environment is
only useful for REAL chroots. With fakechroot, a call to a binary inside
the "chroot" is like a call to any other foreign binary on the system
and the binfmt mechanism will use the hosts qemu-user binary to execute
the foreign binary. So this bit was only some historic cruft I just
somehow forgot to remove after introducing fakechroot in my scripts.
Took a long time for me to spot that though... I was just so used to
copy qemu-arm-static there that I never noticed it was not needed at all
anymore...
2. Using qemu with QEMU_LD_PREFIX
---------------------------------
I can confirm that this patch:
http://lists.nongnu.org/archive/html/qemu-devel/2011-07/msg00460.html
will allow the following:
$ QEMU_LD_PREFIX=/tmp/debian-sid/ ./polystrap.sh -d /tmp/debian-sid/ om-gta02
And my /etc/qemu-binfmt/ directory is completely empty! All libraries
are directly taken from the just extracted rootfs in /tmp/debian-sid/.
Therefor this solution solves all the problems pointed out by Yann
Dirson here:
http://lists.debian.org/debian-embedded/2011/06/msg00104.html
3. tzdata
---------
Turned out, that since tzdata 2011h-1 it is no longer possible to only
supply the debconf values that are actually set for tzdata but all
values have to be supplied:
http://bugs.debian.org/cgi-bin/bugreport.cgi?bug=632720
The default debconfseed.txt files for all targets are now updated to
reflect this new requirement.
have fun!
cheers, josch
Reply to: