| Commit message (Collapse) | Author | Age | Files | Lines |
|
|
|
| |
(cherry picked from commit 29925244c1be2d1d5f71d46eba205278624a1366)
|
|
|
|
|
|
|
|
| |
This is an update to commit 41f25ab3df5f306ac717fa7a6efd58328d30c1ae.
Internal functions should be named guestfs___* (3 underscores) to
avoid clashing with the implementation of actions (2 underscores).
(cherry picked from commit 737181bcd7b1de8c3a613d6282030c34efa78fb6)
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
We split the library code into these separate files:
- guestfs.c: creating handles, closing handles, handle-related variables
- actions.c: generated library-side stubs for each action
- bindtests.c: generated code to test bindings
- launch.c: launching the appliance
- proto.c: the library side of the daemon communications protocol
This is just code movement.
Cherry picked from commit 41f25ab3df5f306ac717fa7a6efd58328d30c1ae
and rebased for stable branch.
|
|
|
|
|
|
| |
Rename these two generated files, in preparation for splitting
up the main src/guestfs.c file.
(cherry picked from commit a617f521a4f695a63ced1c734128dc6c05b7024f)
|
|
|
|
| |
(cherry picked from commit 2fd8c259d3daa88b0cdf98090bb57f3dbd178432)
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
Adding the readonly=on option is not so clever. This causes
qemu to present the disk as read-only to the guest. (The
expected behaviour of snapshots=on,readonly=on was that it
would open the disk O_RDONLY but present a writable disk to
the guest).
Since the guest sees a read-only disk, we are unable to do any
recovery if a filesystem on the disk is inconsistent. This basically
prevents most accesses to live disk images.
What we really want is a qemu option which presents a writable
disk to the guest, but only opens the disk on the host side with
O_RDONLY, to alleviate the udev bug RHBZ#571714.
This reverts commit 676462684e05dd8341dd695762dd99a87d8ec022.
(cherry picked from commit 799d52be4f08f6c70c0e8ba1aa7367ba4cdd78c4)
|
|
|
|
|
|
| |
The 'name' parameter is not used on the right hand side of the
match, so it can be removed.
(cherry picked from commit 2e7da2a2f3bbc6d6db148d7dc2ce238bf56f34db)
|
|
|
|
|
|
|
|
|
|
| |
This resolves a warning from gcc 4.5:
assuming signed overflow does not occur when simplifying
conditional to constant
This page explains the issues in some detail:
http://www.airs.com/blog/archives/120
(cherry picked from commit 321ca1ef91a90cec5b94058b84420e8018e3f1d8)
|
|
|
|
|
|
| |
Only one function currently uses DeviceList. The generated code
unfortunately hard-coded the argument name from that function.
(cherry picked from commit aac51942aab63a9355ad6724345ea923148bf2a9)
|
|
|
|
| |
(cherry picked from commit 5b77be72bc4e46c7a53a24f1eb4cbd107a708f17)
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
There's a thread safety issue with the current OCaml bindings which
is well explained in the bug report:
https://bugzilla.redhat.com/show_bug.cgi?id=604691
This commit fixes the safety issue by copying strings temporarily
before releasing the thread lock. Updated code looks like this:
char *filename = guestfs_safe_strdup (g, String_val (filenamev));
int r;
caml_enter_blocking_section ();
r = guestfs_add_drive_ro (g, filename);
caml_leave_blocking_section ();
free (filename);
if (r == -1)
ocaml_guestfs_raise_error (g, "add_drive_ro");
Also included is a regression test.
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
This add an optional explicit $g->close method which may be
used to force the handle to be closed immediately. Note the
provisos about this method in the manual page entry. Callers
should *not* normally use this method.
The implementation of the handle also changes. Before, the
handle was a blessed reference to an integer (the integer
being the pointer to the C guestfs_h handle). Now we change
this to a hashref containing currently the following field:
_g => pointer to C guestfs_h handle (as an integer)
If this field is not present, it means that the handle has been
explicitly closed. This avoids double-freeing the handle.
The user may add their own fields to this hash in order to store
per-handle data. However any fields whose names begin with
an underscore are reserved for use by the Perl bindings.
This commit also adds a regression test.
This commit also changes the existing warning when you call
a method without a Sys::Guestfs handle as the first parameter,
into an error. This is because such cases are always errors.
|
|
|
|
| |
Read the note in the man page before using this feature.
|
| |
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
Previous commit 4df593496e116dfb635731c058b7627e81fc179c broke the
"file" command on logical volume paths, since these are symbolic
links. We *should* follow these (only).
This inadvertantly broke virt-inspector too, which indicates that
we need more regression testing in this area. Since carrying whole
Fedora images around could make the distribution even larger than
now, I'm not sure at the moment how to do this.
Thanks to Matt Booth for diagnosing this bug.
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
The file call can hang if called on char devices (because we are
using the file -s option).
This is hard to solve cleanly without adding another file API.
However this restricts file to regular files, unless called explicitly
with a /dev/ path. For non-regular files, it will now return a
string like "directory".
There is a small semantic change for symbolic links. Previously
it would not have worked at all on absolute links (or rather, the
results would have been undefined). It would have treated relative
symlinks to regular files as the regular file itself. Now it will
return the string "symbolic link" in both cases.
This commit also makes the API safe when called on untrusted
filesystems. Previously a filesystem might have been set up so
that (eg) /etc/redhat-release was a char device, which would have
caused virt-inspector and virt-v2v to hang. Now it will not hang.
|
| |
|
|
|
|
|
|
|
| |
(RHBZ#599464).
This also adds a regression test for VFAT and (conditionally)
NTFS filesystems.
|
|
|
|
|
|
|
|
| |
Using IfAvailable "featurename" we allow individual tests to
only run if the feature is available in the daemon.
This will allow us to extend testing to a lot more optional
features such as NTFS.
|
|
|
|
| |
This commit is just code motion.
|
| |
|
| |
|
| |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
With this change, the exit status indicates error for non-existent
commands.
$ guestfish -h foo
foo: command not known, use -h to list all commands
$ echo $?
1
$ guestfish help foo
foo: command not known, use -h to list all commands
$ echo $?
1
|
|
|
|
| |
This also adds a regression test.
|
|
|
|
|
|
|
|
|
| |
Fix these calls (see description in RHBZ#597112), but also
deprecate them since the new calls vfs_label and vfs_uuid can
work on any filesystem type.
This also adds a regression test for the original bug reported
in RHBZ#597112.
|
|
|
|
|
|
|
|
|
|
|
| |
These APIs generalize the existing 'get-e2label' and 'get-e2uuid'
calls, to provide calls which should be able to get the label
and UUID for most filesystem types. These use 'blkid' to do the
work.
I have tested that the blkid commands themselves work on RHEL 5.
(Suggested by Yufang Zhang).
|
| |
|
| |
|
| |
|
| |
|
| |
|
| |
|
| |
|
|
|
|
|
|
| |
guestfs_fallocate takes an integer for the length, effectively
limiting it to creating 1GB files. This new call takes an int64_t
for the length, but is otherwise identical.
|
| |
|
| |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
This checks all available optional groups and prints out which
ones are supported by the daemon. Note you must launch the appliance
first.
Example:
><fs> supported
augeas yes
inotify yes
linuxfsuuid yes
linuxmodules yes
linuxxattrs yes
lvm2 yes
mknod yes
ntfs3g yes
ntfsprogs yes
realpath yes
scrub yes
selinux yes
xz yes
zerofree yes
|
| |
|
|
|
|
|
| |
Remove reference to 'ELF weak linking tricks' and replace
with suggestion to use dl* functions.
|
|
|
|
|
| |
See:
http://www.python.org/dev/peps/pep-0353/#conversion-guidelines
|
| |
|
| |
|
| |
|
| |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
This small change uses the gnulib xstrtoll functionality to
enable suffixes on integer parameters in guestfish. For example:
truncate-size /file 1G
(previously you would have had to given the full number).
This also applies to the 'alloc' and 'sparse' commands (and
indirectly to the -N option). The specification for these commands
has changed slightly, in that 'alloc foo 1MB' would now use SI
units, allocating 1000000 bytes instead of a true megabyte. All
existing uses would use 'alloc foo 1M' which still allocates true
megabytes.
|
|
|
|
|
|
|
| |
On Ubuntu <= Karmic, xz-utils was not packaged, and therefore
any xz-related tests would fail. Thus make this an optional
group so that we can test for this and avoid running the tests
if xz utils are not present.
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
This adds additional tests to check that several types of parameter
including String are not NULL when passed to the C functions.
Previously this would cause a segfault inside libguestfs. With
this change, you get an error message / exception.
Of the possible pointer parameters, only OptString is now permitted
to be NULL.
This change does not affect the Perl bindings. This is because Perl
XS code was already adding similar checks if you passed undef into
a parameter expecting a string.
|
|
|
|
|
| |
I haven't checked the list of functions exhaustively, but
these are the obvious ones.
|
| |
|