| Commit message (Collapse) | Author | Age | Files | Lines |
|---|
| | |
|
| | |
|
| | |
|
| |
|
|
|
| |
These APIs allow you to efficiently write and read parts of
files or devices.
|
| |
|
|
|
|
|
|
|
| |
Clear the in-memory history before exiting. This removes
some but not all memory leaks associated with using the GNU
History library. As far as I can tell it is not possible to
free up everything used by GNU History.
(Found by valgrind).
|
| |
|
|
|
|
|
|
|
|
| |
Previously the list of -a, -d, -m, -N parameters were leaked. This
change frees them explicitly.
This is not such an important fix since guestfish is a one-shot
program, but it aids in finding other leaks in future.
(Found by valgrind).
|
| |
|
|
|
|
|
| |
The compiled PCRE regexps used for inspection were being leaked when
the library was unloaded.
(Found by valgrind).
|
| |
|
|
|
|
|
| |
This frees the string containing the name of the appliance
which was previously being leaked during launch.
(Found by valgrind).
|
| | |
|
| |
|
|
|
|
|
|
|
|
| |
The appliance was being completely rebuilt every time guestfsd was updated. This
was because make.sh depended on guestfsd, which it had to do because it
called update.sh to install guestfsd.
This fix removes the call to update.sh in make.sh, and therefore the dependency
on guestfsd. The Makefile already includes a rule to run update.sh when guestfsd
is updated, so this was unnecessary.
|
| |
|
|
|
| |
See also:
http://catless.ncl.ac.uk/Risks/26.17.html#subj13.3
|
| | |
|
| |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
When writing to a RHEV target, virt-v2v launches the libguestfs
appliance with euid:egid = 36:36, which is required to write to
an NFS target using root_squash.
Since we changed to using a cached appliance, this causes an error on
start up, as the cached files are owned by root, but the cache directory
is owned by 36:36. The reason is that bash resets euid to uid and
egid to gid so when febootstrap-supermin-helper is executed, it runs as
root:root. The cache directory was created by libguestfs directly so
it has the correct ownership.
This patch fixes the issue by using explicit fork/exec instead of
system (ie. not going via a shell) and by setting the real UID and
GID to the effective UID and GID before execing.
|
| | |
|
| |
|
|
|
|
| |
The guestfish-only commands such as 'alloc' and 'edit' are
now generated from one place in the generator instead of being
spread around ad-hoc in the C code.
|
| | |
|
| |
|
|
|
| |
This corrects commit b5c287bcd456bdb02d8ec0443483df34f4fd6b5d
and commit 639ca1828b167bf59353f0cd3c8c79c6289bbd5d.
|
| | |
|
| | |
|
| | |
|
| | |
|
| | |
|
| |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
This API is a simpler replacement for the guestfish commands
list-devices / list-partitions / lvs, in the case where you are
just examining a guest by hand to see what it contains.
Typical usage and output in guestfish is like this:
$ guestfish --ro -a /dev/vg_trick/F13x64
><fs> run
><fs> list-filesystems
/dev/vda1: ext4
/dev/vg_f13x64/lv_root: ext4
/dev/vg_f13x64/lv_swap: swap
It can also be used to replace programs that try to mount
devices to determine if they are mountable filesystems.
|
| |
|
|
|
| |
This adds a formal API for going from a partition to the containing
device, eg. /dev/sda1 -> /dev/sda
|
| |
|
|
|
|
| |
This is for testing functions that return a device or partition
name, so that we can compare the return value with the canonical
device name (eg. "/dev/vda1" == "/dev/sda1").
|
| | |
|
| | |
|
| | |
|
| | |
|
| | |
|
| | |
|
| |
|
|
|
|
| |
All other vmchannel methods are obsolete, but we were still trying
to check for them. This replaces all of them with a simple check
for virtio-serial.
|
| | |
|
| | |
|
| | |
|
| | |
|
| |
|
|
|
|
| |
This is a workaround until
https://bugzilla.redhat.com/show_bug.cgi?id=630583
is fixed (bug in Linux 2.6.36).
|
| | |
|
| |
|
|
|
| |
It seems that linux-image (ie. the kernel) is omitted in some
versions of the base packages.
|
| |
|
|
|
| |
If it's a symlink then the succeeding mount commnd will fail.
Remove it and make a /proc directory.
|
| | |
|
| |
|
|
|
| |
hivex library has been required since we moved the inspection
code to C. Check for this in configure.ac.
|
| | |
|
| |
|
|
|
|
| |
Augeas has been required since we moved the inspection code to C,
however we were not correctly enforcing this in configure.ac, nor
correctly linking to the library until now.
|
| | |
|
| | |
|
| |
|
|
|
|
|
|
|
| |
Because this used to be compiled into the C test, it changed
every time the ISO was rebuilt (which because of Makefile deps
was every run).
Now it is calculated at runtime so the C test file doesn't keep
changing.
|
| |
|
|
|
| |
This was one reason why capitests/tests.c changed every time the
generator was run.
|
| |
|
|
|
|
|
|
|
|
| |
'src/generator.ml' is no more. Instead the generator is logically
split up over many different source files.
Read generator/README for help and tips.
We compile the generator down to bytecode, not native code. This
means it will run more slowly, but is done for maximum portability.
|
| | |
|
