Just code motion.
This commit makes it clearer what is a utility and what is part of the
library. It also makes it clear that we should rename:
guestfs-internal-frontend.h -> utils.h
guestfs-internal-frontend-cleanups.h -> cleanups.h (?)
but this commit does not make that change.
This commit, which is just code motion, moves the common XDR protocol
code (libprotocol) and the common errno handling (liberrnostring) into
libraries which are each built once and shared between the library and
daemon.
The error is:
gobject/docs/gtk-doc.make:52: warning: CLEANFILES multiply defined in condition TRUE ...
gobject/docs/Makefile.am:95: 'gobject/docs/gtk-doc.make' included from here
common-rules.mk:28: ... 'CLEANFILES' previously defined here
gobject/docs/Makefile.am:18: 'subdir-rules.mk' included from here
subdir-rules.mk:20: 'common-rules.mk' included from here
In the ancient version of PCRE in RHEL 5, partial match support does
not support /.*/, returning PCRE_ERROR_BADPARTIAL. However the
equivalent regular expression /(?:.)*/ works fine, so use that
instead.
RHEL 5 has ancient Gtk 2.10 which lacks several functions and whole
features we need (eg the spinner). Add various macros and function
definitions to work around this.
The guestfs_canonical_device_name API was rewriting /dev/mdX as
/dev/sdX. This is wrong since (eg) /dev/sd0 is not a device name, so
if you pass the canonicalized name back to the API it will fail.
virt-v2v was one tool doing this.
Bug found by valgrind on aarch64:
==21350== Syscall param faccessat(pathname) points to unaddressable byte(s)
==21350== at 0x4F05274: access (access.c:29)
==21350== by 0x4954DC3: guestfs_int_get_uefi (appliance-uefi.c:90)
==21350== by 0x49752D7: launch_libvirt (launch-libvirt.c:413)
==21350== by 0x496FD83: guestfs_impl_launch (launch.c:98)
==21350== by 0x4902003: guestfs_launch (actions-3.c:145)
[...]
==21350== Address 0x0 is not stack'd, malloc'd or (recently) free'd
Because we used 'return' directly, the effect of the CAMLparam0()
macro was not undone, resulting in local roots corruption when the
visit callback throws an exception.
The visit_tests unit test picked this up, but the corruption was only
seen on aarch64.
Fixes commit 4a62e52111.
appliance-uefi.c: In function 'guestfs_int_get_uefi':
appliance-uefi.c:67:9: error: implicit declaration of function 'access' [-Werror=implicit-function-declaration]
if (access (codefile, R_OK) == 0 && access (varsfile, R_OK) == 0) {
^~~~~~
appliance-uefi.c:67:5: error: nested extern declaration of 'access' [-Werror=nested-externs]
if (access (codefile, R_OK) == 0 && access (varsfile, R_OK) == 0) {
^~
appliance-uefi.c:67:27: error: 'R_OK' undeclared (first use in this function)
if (access (codefile, R_OK) == 0 && access (varsfile, R_OK) == 0) {
^~~~
appliance-uefi.c:67:27: note: each undeclared identifier is reported only once for each function it appears in
cc1: all warnings being treated as errors
Fixes commit dd519e8a8e.
Run the following command over the source:
perl -pi.bak -e 's/(20[01][0-9])-2016/$1-2017/g' `git ls-files`
(Thanks Rich for the perl snippet, as used in past years.)
EDD (https://en.wikipedia.org/wiki/Enhanced_Disk_Drive) is a BIOS INT 13h
subcall for communicating the extended size of the boot disk over 8GB
to the OS. Since libguestfs doesn't use a boot disk, and the
appliance disk is limited to 4GB, and we use virtio-scsi, this is all
useless.
EDD is also broken currently on RHEL 7.3, see:
https://bugzilla.redhat.com/show_bug.cgi?id=1404287
Also the EDD probing takes significant extra time (about 80ms on my
laptop), and using edd=off reduces this time although doesn't entirely
eliminate it.
One of the tests for btrfs_subvolume_show tries to invoke it on the
toplevel subvolume, expecting its failure (since btrfs-progs used to not
be able to list that). However, since v4.8.3 btrfs-progs can do that
(even if the data returned for that subvolume is limited/incomplete),
and thus this test fails because the API does not fail anymore.
Since we can consider that functionality working (for some kind of it),
then drop this test.
