'make extra-tests' was a monolithic set of tests that did all sorts of
things: valgrind, tests over local guests, tests with upstream qemu,
tests with upstream libvirt, tests with the appliance attach method.
This made it hard to perform individual tests, eg. just valgrind
testing. It was also hard to maintain because the tests were not
located in the same directories as the programs and sometimes
duplicated tests that were run elsewhere.
This commit splits up 'make extra-tests' into 5 separate targets:
make check-valgrind # run a subset of tests under valgrind
make check-valgrind-local-guests # test under valgrind with local guests
make check-with-appliance # test with attach-method == appliance
make check-with-upstream-qemu # test with an alternate/upstream qemu
make check-with-upstream-libvirt # test with an alternate/upstream libvirt
(You can also still run 'make extra-tests' which is now simply
a rule that runs the above 5 targets in order).
This replaces everything that was in the tests/extra directory,
so that has now gone.
This command was not tested at all. As a result we didn't notice that
it was broken for a long time (RHBZ#853159).
This adds a test that drives the command through a pty. It uses the
perl 'Expect' module, although this is not required.
This uses Pod::Simple so it properly parses the input POD and can
generate proper custom output as required specifically by libguestfs.
One immediate benefit is that links between and within manual pages
now work mostly correctly.
Earlier versions of qemu contained a bug in the qcow2 code which
causes qemu to segfault when shutting down and flushing its internal
cache, and this can result in data loss.
QEMU 1.0 was released at the end of 2011.
Remove all the cruft about detecting broken -machine type which
was only required for QEMU 0.15.
This also reverts commit 30ecbf3ec2.
Even on ARM you can pass -machine accel=kvm:tcg and qemu does the
right thing, so I'm not sure why we wanted to disable that.
Note that this support is optional: To enable it, install the
ocaml-gettext library from
http://forge.ocamlcore.org/projects/ocaml-gettext . If this library
is not installed, then configure detects this and inserts dummy
gettext functions that do nothing.
Netpbm and icoutils (wrestool) have always been dependencies. Since
they are not always present, make these into optional dependencies
(which they were, sort of, before).
Also document these dependencies in the README file.
This switches virt-sysprep to use guestmount instead of guestfish.
This makes the script a little bit easier to modify for sysadmins.
This commit also adds:
- dhcp-client-state
- dhcp-server-state
- logfiles
- random-seed
- smolt-uuid
- yum-uuid
SELinux relabelling, and a section on security in the manual page.
Update and verify the list of requirements, by checking it against
both configure.ac and the Fedora specfile.
Remove some obsolete sections that covered historical ground.
Remove the hack that let you run ./fish/guestfish or
./fuse/guestmount. You now have to do:
./run ./fish/guestfish
or
./run ./fuse/guestmount
to run these programs without installing.
This is a fairly straightforward translation of Perl virt-resize into
OCaml. It is bug-for-bug and feature-for-feature identical to the
Perl version, except as noted below.
The motivation is to have a more solid, high-level, statically safe
compiled language to go forwards with fixing some of the harder bugs
in virt-resize. In particular contracts between different parts of
the program are now handled by statically typed structures checked at
compile time, instead of the very ad-hoc unchecked hash tables used by
the Perl version.
OCaml and the ocaml-pcre library (Perl-Compatible Regular Expressions
bindings for OCaml) are required.
Extra features in this version:
- 32 bit hosts are now supported.
- We try hard to handle the case where the target disk is not "clean"
(ie. all zeroes). It usually works for this case, whereas the
previous version would usually fail. However it is still
recommended that the system administrator creates a fresh blank disk
for the target before running the program.
- User messages are a bit more verbose and helpful. You can turn
these off with the -q (--quiet) option.
There is one lost feature:
- Ability to specify >= T (terabytes) sizes in command line size
expressions has been removed. This probably didn't work in the Perl
version.
Other differences:
- The first partition on the target is no longer aligned; instead we
place it at the same sector as on the source. I suspect that
aligning it was causing the bootloader failures.
- Because it's easier, we do more sanity checking on the source disk.
This might lead to more failures, but they'd be failures you'd want
to know about.
- The order in which operations are performed has been changed to make
it more logical. The user should not notice any functional
difference, but debug messages will be quite a bit different.
- virt-resize is a compiled binary, not a script.
This removes all support for building the ordinary / old
style appliance using febootstrap 2.x, debootstrap, debirf,
fakeroot and fakechroot.
Instead this uses febootstrap 3.x to build the supermin appliance
in a simpler cross-distro manner.
At some point we removed the last thing that required
xml-light, but were still testing for it at various places
in the build. This removes all traces.
This also makes libxml2 and libvirt into optional dependencies.
If they are missing then the core API will print an error, as
will the '-d' option to guestfish.
This change makes these libraries optional. If they are not
available at compile time then certain core API features will
be disabled (see below).
This also changes PCRE detection to use pkg-config instead
of the ad hoc autoconf checks.
The large inspect.c file has been split out into separate
function-specific files.
file-architecture: requires pcre & libmagic
inspection: requires pcre & hivex
If either the daemon sends back an errno, or a system call
fails in the library, save the errno in the handle and then
make it available to callers through the guestfs_last_errno
function.
Use febootstrap-supermin-helper's new -u and -g command line options to setuid,
rather than doing it in libguestfs.
This resolves an issue with the generation of the cached appliance checksum. The
checksum was being generated by a call to febootstrap-supermin-helper through
popen(). Unfortunately, a bash misfeature meant that euid would be reset to uid,
and the checksum was generated for uid, not euid. When virt-v2v is writing to a
RHEV target, uid == 0 and euid == 36, which resulted in a cached appliance being
created for root with permissions for uid 36.
Note this requires febootstrap 2.10.
