Jim Meyering [Thu, 6 Aug 2009 13:25:03 +0000 (15:25 +0200)]
define ATTRIBUTE_UNUSED
* daemon/daemon.h (__attribute__, ATTRIBUTE_UNUSED): Define.
Jim Meyering [Fri, 7 Aug 2009 18:07:03 +0000 (20:07 +0200)]
daemon/file.c: remove duplicate absolute-path check
* daemon/file.c (do_file): Remove redundant use of ABS_PATH.
It's redundant because the preceding line invokes NEED_ROOT_OR_IS_DEVICE,
which also invokes ABS_PATH.
Richard Jones [Fri, 7 Aug 2009 16:38:44 +0000 (17:38 +0100)]
Tidy up generation of java/Makefile.inc.
(Just whitespace and comment changes, and small restructuring
of the code).
Richard Jones [Fri, 7 Aug 2009 08:51:04 +0000 (09:51 +0100)]
Debug: Improve the way the qemu command line is printed.
Change the way the qemu command is displayed to look like
this:
/usr/bin/qemu-kvm \
-drive file=/tmp/test.img,cache=off,if=virtio \
-m 500 \
-no-reboot \
-kernel /tmp/libguestfsHBJHRh/kernel \
-initrd /tmp/libguestfsHBJHRh/initrd \
[...]
This allows the command line to be copied and pasted directly
into the shell, and also makes it simpler to read.
Richard Jones [Fri, 7 Aug 2009 08:31:35 +0000 (09:31 +0100)]
Allow network interface to be configured.
Add ./configure --with-net-if=(virtio|ne2k_pci) option.
This lets you workaround the following virtio_net bug:
https://bugzilla.redhat.com/show_bug.cgi?id=516022
Richard Jones [Fri, 7 Aug 2009 07:32:56 +0000 (08:32 +0100)]
Tests: swapon_device test mkswap /dev/sda1 instead of /dev/sdb.
On RHEL 5, mkswap /dev/sdb fails, even with the '-f' option
(contradictory to what the manual page says).
This commit adds a new 'InitPartition' initialization which
just creates /dev/sda1, and then does the swap test on that.
Richard Jones [Thu, 6 Aug 2009 17:46:57 +0000 (18:46 +0100)]
Always pass mkswap -f parameter.
Otherwise mkswap will give a silly error if you ask it to
swap on a whole device.
Richard Jones [Thu, 6 Aug 2009 17:40:59 +0000 (18:40 +0100)]
Tests: Don't test block size in statvfs test.
The block size changes between kernel releases, causing
a false failure in the tests. The error was:
test_statvfs_0: bsize was 65536, expected 131072
Richard Jones [Thu, 6 Aug 2009 16:02:25 +0000 (17:02 +0100)]
Version 1.0.66
Richard Jones [Thu, 6 Aug 2009 14:44:20 +0000 (15:44 +0100)]
RHEL 5: inotify_init1 call did not exist on RHEL 5.
Richard Jones [Thu, 6 Aug 2009 13:35:46 +0000 (14:35 +0100)]
Fix: segfault in tab completion (RHBZ#516024).
Actually this fixes two bugs: 'strs' was not being freed on every
path, and the tab completion segfault described in the bug report.
Matthew Booth [Thu, 6 Aug 2009 13:14:18 +0000 (14:14 +0100)]
Fix errno check in readdir in devsparts.c
Richard Jones [Thu, 6 Aug 2009 12:38:17 +0000 (13:38 +0100)]
Add ne2k-pci driver to the appliance.
This allows people to use the ne2k-pci network driver, for those
situations where the virtio_net driver is broken (cough 516022 cough).
Richard Jones [Thu, 6 Aug 2009 11:52:49 +0000 (12:52 +0100)]
Make user network numbering explicit.
Use:
-net user,vlan=0,net=10.0.2.0/8
just to make the (already implicit) 10.0.2.x network explicit.
Richard Jones [Thu, 6 Aug 2009 11:10:52 +0000 (12:10 +0100)]
appliance: More debug added to /init.
Matthew Booth [Thu, 6 Aug 2009 12:07:51 +0000 (13:07 +0100)]
Exclude daemon/lib and daemon/tests from being added to po/POTFILES.in
Jim Meyering [Thu, 6 Aug 2009 11:50:11 +0000 (07:50 -0400)]
build: remove bootstrap's --gnulib-srcdir option
...because it probably didn't work, and even if it did, we've
discovered that using a separate git repo like that can lead
to subtle mix-ups.
Also, fix invocation of gnulib-tool in daemon/.
Jim Meyering [Thu, 6 Aug 2009 11:47:48 +0000 (13:47 +0200)]
build: fix build failure
* bootstrap (gnulib_tool): Create lib and tests directories
under daemon/ before running gnulib-tool there.
Jim Meyering [Mon, 3 Aug 2009 10:44:51 +0000 (11:44 +0100)]
daemon: use gnulib
* daemon/Makefile.am (SUBDIRS): Define.
(AM_CPPFLAGS): Define, to include from gnulib's lib/
(LDADD): Define, to link with gnulib's libgnu.a.
* daemon/configure.ac: Use AC_CONFIG_AUX_DIR([build-aux]),
gl_EARLY and gl_INIT.
(AC_CONFIG_FILES): Add lib/Makefile and tests/Makefile
* daemon/m4/gnulib-cache.m4: New file, generated by running
../.gnulib/gnulib-tool --import --with-tests hash
* daemon/.gitignore: Ignore all of the imported files.
build: tell bootstrap about daemon/
* bootstrap: Run gnulib-tool --update in daemon/.
Remove bootstrap's --gnulib-srcdir option, because it probably
didn't work, and even if it did, we've discovered that using
a separate git repo like that can lead to subtle mix-ups.
Richard Jones [Thu, 6 Aug 2009 10:01:20 +0000 (11:01 +0100)]
appliance: Include more debug output in the /init script.
Output more debugging information from this script, to enhance the
usefulness of LIBGUESTFS_DEBUG output.
Matthew Booth [Mon, 3 Aug 2009 15:04:50 +0000 (16:04 +0100)]
Recognise cd-rom devices in devsparts.c
Also:
* Un-duplicate device detection code by creating a common mapping function.
* Add some more comments.
Matthew Booth [Wed, 5 Aug 2009 22:33:22 +0000 (23:33 +0100)]
Fix test failures in upload and download (RHBZ#515764)
The tests hardcoded an MD5 of the test file, COPYING.LIB, whose contents had
change. This change causes it to compute the MD5 rather than hardcoding it.
OCaml provided by rjones.
Richard Jones [Thu, 6 Aug 2009 07:56:39 +0000 (08:56 +0100)]
Add copyright and license notice to regressions/test-noexec-stack.pl
Matthew Booth [Wed, 5 Aug 2009 16:32:02 +0000 (17:32 +0100)]
Add some newly-untracked files to .gitignore
Matthew Booth [Mon, 3 Aug 2009 19:29:24 +0000 (20:29 +0100)]
Fix swapon_label test
The swapon label test has a side-effect, in that it creates a new partition.
This causes the later failure of the list_partitions test.
This change causes the swapon_label test to zero /dev/sdb after the test, and
then re-read the partition table.
Matthew Booth [Wed, 5 Aug 2009 15:00:12 +0000 (16:00 +0100)]
Add resultant change to po/POTFILES.in
I missed this when committing
7e9cb884492aec243337ffc8e4432a9ff2690956
Matthew Booth [Wed, 5 Aug 2009 13:55:54 +0000 (14:55 +0100)]
Add a test for an executable stack in resultant binaries
Jim Meyering [Wed, 5 Aug 2009 12:01:38 +0000 (08:01 -0400)]
build: generate some just-removed files in po/
* bootstrap: Generate po/Makevars and po/LINGUAS.
Jim Meyering [Tue, 4 Aug 2009 12:21:32 +0000 (14:21 +0200)]
build: remove trailing blank lines; enable the syntax-check rule
* cfg.mk (disable_temporarily): Remove sc_prohibit_trailing_blank_lines.
* appliance/Makefile.am: Remove trailing blank line(s).
* appliance/debian/modules/y0_install-guestfsd: Likewise.
* appliance/make.sh.in: Likewise.
* appliance/packagelist.in: Likewise.
* appliance/update.sh.in: Likewise.
* haskell/run-bindtests: Likewise.
* ocaml/run-bindtests: Likewise.
* python/run-python-tests: Likewise.
* recipes/squashfs.example: Likewise.
* ruby/run-ruby-tests: Likewise.
Jim Meyering [Tue, 4 Aug 2009 12:19:29 +0000 (14:19 +0200)]
build: exempt symlinks from trailing-blank-line prohibition
* cfg.mk (sc_prohibit_trailing_blank_lines): Exempt symlinks.
Jim Meyering [Tue, 4 Aug 2009 12:03:17 +0000 (14:03 +0200)]
build: enable syntax-check: TAB-only indentation in Makefiles
* cfg.mk (disable_temporarily): Remove sc_makefile_TAB_only_indentation.
Jim Meyering [Tue, 4 Aug 2009 12:01:38 +0000 (14:01 +0200)]
build: tweak HACKING so we pass "make syntax-check"
* HACKING: Convert leading TABs.
Jim Meyering [Tue, 4 Aug 2009 12:00:11 +0000 (14:00 +0200)]
build: remove trailing blanks, enable syntax-check
* regressions/test-cancellation-download-librarycancels.sh:
* src/generator.ml:
* cfg.mk (disable_temporarily): Remove sc_trailing_blank.
* .x-sc_trailing_blank: New file.
Jim Meyering [Tue, 4 Aug 2009 12:46:38 +0000 (14:46 +0200)]
maint: use COPYING.LIB version 2.1
* COPYING.LIB: Update to 2.1.
* .x-sc_TAB_in_indentation: Also exempt COPYING.LIB.
Jim Meyering [Tue, 4 Aug 2009 11:55:51 +0000 (13:55 +0200)]
build: don't emit trailing blanks, remove generated file
* ocaml/Makefile.am (.depend): Don't redirect directly to $@.
Filter out trailing blanks.
* ocaml/.depend: Regenerate.
Jim Meyering [Tue, 4 Aug 2009 11:49:19 +0000 (13:49 +0200)]
build: remove more files added by ./autogen.sh
* po/LINGUAS: Remove file.
* po/Makefile.in.in: Likewise.
* po/Makevars: Likewise.
* po/Rules-quot: Likewise.
* po/boldquot.sed: Likewise.
* po/en@boldquot.header: Likewise.
* po/en@quot.header: Likewise.
* po/insert-header.sin: Likewise.
* po/quot.sed: Likewise.
* po/remove-potcdate.sin: Likewise.
Jim Meyering [Tue, 4 Aug 2009 08:47:17 +0000 (04:47 -0400)]
build: tell aclocal to search gnulib/m4, too
* Makefile.am (ACLOCAL_AMFLAGS): Append "-I gnulib/m4".
Jim Meyering [Mon, 3 Aug 2009 14:55:34 +0000 (10:55 -0400)]
Document and enforce the new spaces-only indentation policy.
* cfg.mk (disable_temporarily): Remove sc_TAB_in_indentation,
to enable the "make syntax-check" that enforces this.
* .x-sc_TAB_in_indentation: New file. Exceptions.
* HACKING: Document the policy, with tips on how to conform.
Jim Meyering [Mon, 3 Aug 2009 15:10:13 +0000 (11:10 -0400)]
Convert all TABs-as-indentation to spaces.
Do it by running this command:
[exempted files are matched via .x-sc_TAB_in_indentation]
git ls-files \
| pcregrep -vf .x-sc_TAB_in_indentation \
| xargs pcregrep -l '^ *\t' \
| xargs perl -MText::Tabs -ni -le \
'$m=/^( *\t[ \t]*)(.*)/; print $m ? expand($1) . $2 : $_'
Jim Meyering [Mon, 3 Aug 2009 15:00:22 +0000 (11:00 -0400)]
Remove files imported via autogen.sh.
* m4/.gitignore: Ignore these files.
* m4/codeset.m4: Remove file.
* m4/gettext.m4: Likewise.
* m4/glibc2.m4: Likewise.
* m4/glibc21.m4: Likewise.
* m4/iconv.m4: Likewise.
* m4/intdiv0.m4: Likewise.
* m4/intl.m4: Likewise.
* m4/intldir.m4: Likewise.
* m4/intlmacosx.m4: Likewise.
* m4/intltool.m4: Likewise.
* m4/inttypes-pri.m4: Likewise.
* m4/inttypes.m4: Likewise.
* m4/inttypes_h.m4: Likewise.
* m4/isc-posix.m4: Likewise.
* m4/lcmessage.m4: Likewise.
* m4/lib-ld.m4: Likewise.
* m4/lib-link.m4: Likewise.
* m4/lib-prefix.m4: Likewise.
* m4/lock.m4: Likewise.
* m4/longdouble.m4: Likewise.
* m4/longlong.m4: Likewise.
* m4/nls.m4: Likewise.
* m4/po.m4: Likewise.
* m4/printf-posix.m4: Likewise.
* m4/progtest.m4: Likewise.
* m4/signed.m4: Likewise.
* m4/size_max.m4: Likewise.
* m4/stdint_h.m4: Likewise.
* m4/uintmax_t.m4: Likewise.
* m4/ulonglong.m4: Likewise.
* m4/visibility.m4: Likewise.
* m4/wchar_t.m4: Likewise.
* m4/wint_t.m4: Likewise.
* m4/xsize.m4: Likewise.
Jim Meyering [Mon, 3 Aug 2009 15:09:13 +0000 (11:09 -0400)]
allow use of #if HAVE_CONFIG_H, for example code
* .x-sc_prohibit_have_config_h: New file, to exempt examples.
Jim Meyering [Mon, 3 Aug 2009 13:39:09 +0000 (15:39 +0200)]
guestfs: fix typo in my recent change
* src/guestfs.c (guestfs_perrorf): Rename former err to errnum,
to avoid compilation error.
Jim Meyering [Mon, 3 Aug 2009 13:21:44 +0000 (15:21 +0200)]
Guard #inclusion of config.h in examples.
* examples/hello.c [HAVE_CONFIG_H]: Include <config.h> conditionally.
* examples/to-xml.c: Likewise.
Jim Meyering [Mon, 3 Aug 2009 12:32:44 +0000 (08:32 -0400)]
build: reenable "syntax-check" rule: sc_const_long_option
* cfg.mk (disable_temporarily): Remove sc_const_long_option.
* daemon/guestfsd.c (main): Declare long_options to be "const".
* fish/fish.c (main): Likewise.
* test-tool/test-tool.c (main): Likewise.
Jim Meyering [Mon, 3 Aug 2009 09:53:26 +0000 (05:53 -0400)]
build: temporarily disable failing "syntax-check" rules
* cfg.mk (disable_temporarily): Define.
Jim Meyering [Mon, 3 Aug 2009 09:18:10 +0000 (05:18 -0400)]
maint: use a git submodule for gnulib
* .gitmodules: New file, to track gnulib.
* .gnulib: Submodule directory.
* Makefile.am (EXTRA_DIST): Don't list config.rpath or
gitlog-to-changelog.
* autogen.sh: Adapt to use the new submodule.
* cfg.mk: New file.
(SUBDIRS): Add gnulib/lib and gnulib/tests.
(dist-hook): Reflect new location of getlog-to-changelog.
* configure.ac: Set build-aux/ as AUX_DIR.
Invoke gl_EARLY and gl_INIT.
(AC_CONFIG_FILES): Add gnulib/lib/Makefile and gnulib/tests/Makefile.
Jim Meyering [Mon, 3 Aug 2009 09:33:35 +0000 (05:33 -0400)]
always include <config.h>
* ocaml/guestfs_c.c: Include <config.h>.
* examples/to-xml.c: Likewise.
* examples/hello.c: Likewise.
Jim Meyering [Mon, 3 Aug 2009 10:58:16 +0000 (06:58 -0400)]
build: add -I option to get config.h.
* ocaml/Makefile.am (AM_CPPFLAGS): Define.
(guestfs_c.o, guestfs_c_actions.o): Use it.
Jim Meyering [Mon, 3 Aug 2009 12:43:54 +0000 (08:43 -0400)]
guestfs: don't fault upon failed vasprintf
* src/guestfs.c (guestfs_perrorf): Handle failed vasprintf.
Matthew Booth [Mon, 3 Aug 2009 10:43:38 +0000 (11:43 +0100)]
Cleanup whitespace warnings in Lib.pm
Matthew Booth [Fri, 31 Jul 2009 16:31:54 +0000 (17:31 +0100)]
Use grub entries to find Linux kernels
This change adds grub parsing to Lib.pm. It adds the following structure to $os:
{boot}
->{configs}
->[0]
->{title} = "Fedora (2.6.29.6-213.fc11.i686.PAE)"
->{kernel} = \kernel
->{cmdline} = "ro root=/dev/mapper/vg_mbooth-lv_root rhgb"
->{initrd} = \initrd
->{default} = 0
The kernel and initrd entries are just references to their top level entries
under kernels and initrd_modules respectively.
It also changes the way Linux kernels and initrd are discovered. Instead of
searching /lib/modules and /boot for files with matching names, kernels and
initrds are scanned as they are discovered in grub.conf.
Additionally, the following attributes are added to the kernels top level entry:
* path
The path to the kernel's vmlinuz file.
* package
The name of the package which installed the kernel.
The xml output of virt-inspector is updated to reflect all of the above changes.
Richard W.M. Jones [Fri, 31 Jul 2009 22:24:11 +0000 (23:24 +0100)]
guestfish: Display RStructList results more pleasantly.
Richard W.M. Jones [Fri, 31 Jul 2009 21:31:05 +0000 (22:31 +0100)]
Add interface to Linux 'inotify' API.
Richard W.M. Jones [Fri, 31 Jul 2009 14:57:46 +0000 (15:57 +0100)]
lib: Add selinux=0 to default kernel command line.
SELinux exists in a very disturbed state if it is enabled at
boot time, but no policy is loaded. In particular, it messes
up the security.selinux extended attributes on files in a
not-very-useful way.
We can't enable SELinux because we don't know what policy
can or should be loaded. Therefore it's best to disable it
completely.
Matthew Booth [Fri, 31 Jul 2009 13:30:09 +0000 (14:30 +0100)]
Update incorrect comment in Lib.pm
Related: change the name of the function the comment describes to be more
accurate.
Richard W.M. Jones [Fri, 31 Jul 2009 11:12:29 +0000 (12:12 +0100)]
Improve warnings about missing tests.
Don't warn where a command just has no tests.
Instead check other commands' tests so we get a definitive
(and much smaller) list of commands that are not tested anywhere.
Richard W.M. Jones [Fri, 31 Jul 2009 10:55:38 +0000 (11:55 +0100)]
New commands: swapon-*, swapoff-*, mkswap-file.
swapon-device
swapoff-device
swapon-file
swapoff-file
swapon-label
swapoff-label
swapon-uuid
swapoff-uuid
mkswap-file
Richard W.M. Jones [Fri, 31 Jul 2009 07:57:10 +0000 (08:57 +0100)]
New command: 'fallocate' to (pre-)allocate sized files.
Richard W.M. Jones [Fri, 31 Jul 2009 07:39:43 +0000 (08:39 +0100)]
New commands: 'ln', 'ln-f', 'ln-s', 'ln-sf' and 'readlink'.
These commands can be used to make hard and symbolic links. The
readlink command is used to read existing symbolic links.
Richard W.M. Jones [Fri, 31 Jul 2009 06:39:30 +0000 (07:39 +0100)]
Add 'realpath' command.
Richard W.M. Jones [Fri, 31 Jul 2009 06:40:57 +0000 (07:40 +0100)]
Ignore failure of 'mv' commands in appliance Makefile.
A lot of people report these as 'errors' when they are not.
Hide the error messages.
Jim Meyering [Thu, 30 Jul 2009 12:03:48 +0000 (14:03 +0200)]
configure.ac: more quoting
Also, prohibit an invalid value of $JAVA_HOME,
since it must be used unquoted below.
Richard Jones [Thu, 30 Jul 2009 11:10:52 +0000 (12:10 +0100)]
build: If guestfwd test fails, need to print AC_MSG_RESULT 'no'
Jim Meyering [Thu, 30 Jul 2009 08:50:49 +0000 (10:50 +0200)]
build: fix test for --nocompress option
Richard W.M. Jones wrote:
> On Wed, Jul 29, 2009 at 10:50:44PM +0200, Jim Meyering wrote:
>> The test for febootstrap-to-initramfs' --nocompress option
>> was always failing for me on F11. Here's why:
> [...]
>
> Ouch that's obscure. I've applied this patch and the previous
> one you sent too. Thanks!
Our messages crossed.
Rebasing my fixed patch and adjusting the log:
>From
9e7846da50ceeee57187f703835bd3975e789719 Mon Sep 17 00:00:00 2001
From: Jim Meyering <meyering@redhat.com>
Date: Wed, 29 Jul 2009 16:33:02 -0400
Subject: [PATCH] build: quote the other option test similarly
* configure.ac: Quote properly.
Jim Meyering [Wed, 29 Jul 2009 20:50:44 +0000 (22:50 +0200)]
build: fix test for --nocompress option
The test for febootstrap-to-initramfs' --nocompress option
was always failing for me on F11. Here's why:
$ bash ~/w/co/libguestfs:210-comment-fix
$ t=`febootstrap-to-initramfs 2>&1`
[Exit 1]
$ if ! echo $t|grep -sq -- --nocompress; then echo not found; fi
not found
Notice: without quotes, the [--nocompress] term expands to "k",
because I happen to have a temporary file named "k":
$ echo $t
Usage: febootstrap-to-initramfs [--files=filelist] k DIR Please read
febootstrap-to-initramfs(8) man page for more information.
$ echo "$t"
Usage: febootstrap-to-initramfs [--files=filelist] [--nocompress] DIR
Please read febootstrap-to-initramfs(8) man page for more information.
a simpler example gives a clue:
$ touch a b c
$ echo [--z]
a b c
$ echo [--a]
a
$ echo [--b]
a b
it's interpreting [--nocompress] as a range: "-" through "n", plus
the other characters, ocmpres. "k" falls in the --n range.
Anyhow, here's the fix:
>From
84855642ed41828d01d55123cfab8d8dede759c1 Mon Sep 17 00:00:00 2001
From: Jim Meyering <meyering@redhat.com>
Date: Wed, 29 Jul 2009 16:33:02 -0400
Subject: [PATCH] build: fix test for --nocompress option
* configure.ac: Fix underquoting bug that would cause the test
for febootstrap-to-initramfs' --nocompress option always to fail
with certain-letter-named files in the top directory.
Jim Meyering [Wed, 29 Jul 2009 18:48:45 +0000 (14:48 -0400)]
fix comments; move declarations
* daemon/devsparts.c (do_list_devices, do_list_partitions):
Remove stray words in comments.
Move declarations down to definition.
Richard Jones [Wed, 29 Jul 2009 22:02:42 +0000 (23:02 +0100)]
Whitespace and comment changes.
Richard Jones [Wed, 29 Jul 2009 21:53:23 +0000 (22:53 +0100)]
df: Fix alignment of columns.
Richard Jones [Wed, 29 Jul 2009 20:34:06 +0000 (21:34 +0100)]
guestfish: Make more strings translatable.
However this doesn't yet attempt to translate the POD
command documentation. We need a plan to do that.
Richard Jones [Wed, 29 Jul 2009 17:46:55 +0000 (18:46 +0100)]
tests: Use squashfs for static data where possible.
Instead of creating a new InitBasicFS for just about every test,
where a test doesn't need to write to the filesystem it is far
faster to use prepared data on an InitSquashFS instead.
This commit changes as many tests as possible to make this so,
making the tests much faster.
Richard Jones [Wed, 29 Jul 2009 17:44:46 +0000 (18:44 +0100)]
tests: Fix read_file test.
The test ignored the fact that this function returns a (char *, size_t)
pair, and just treated the char * as a string. This of course would
fail if the string didn't happen to be nul-terminated.
The tests are updated to add a new TestOutputBuffer type which should
be used with functions that return RBufferOut.
Richard Jones [Wed, 29 Jul 2009 17:10:48 +0000 (18:10 +0100)]
Implement '*grep*' family of commands.
Richard Jones [Wed, 29 Jul 2009 15:51:10 +0000 (16:51 +0100)]
Updated Polish translations (Piotr Drąg).
Richard Jones [Wed, 29 Jul 2009 15:09:29 +0000 (16:09 +0100)]
Version 1.0.65.
Richard Jones [Wed, 29 Jul 2009 14:59:28 +0000 (15:59 +0100)]
Updated PO files.
Jim Meyering [Wed, 29 Jul 2009 12:16:03 +0000 (14:16 +0200)]
build: avoid locale-specific changes in generated, VC'd file
When I build with LC_ALL=C in my environment,
the all-local rule generates po/POTFILES.in that
is sorted differently from the on that is checked in:
diff --git a/po/POTFILES.in b/po/POTFILES.in
index ca01b3d..154915a 100644
--- a/po/POTFILES.in
+++ b/po/POTFILES.in
@@ -63,12 +63,11 @@ fish/tilde.c
fish/time.c
inspector/virt-inspector.pl
java/com_redhat_et_libguestfs_GuestFS.c
-ocaml/guestfs_c_actions.c
ocaml/guestfs_c.c
+ocaml/guestfs_c_actions.c
perl/bindtests.pl
-perl/Guestfs.c
-perl/lib/Sys/Guestfs/Lib.pm
perl/lib/Sys/Guestfs.pm
+perl/lib/Sys/Guestfs/Lib.pm
python/guestfs-py.c
ruby/ext/guestfs/_guestfs.c
src/guestfs-actions.c
If we generate that file so that sort always uses the C locale, then,
this type of difference will not arise. Here's the patch to fix the
rule as well as to reflect the change in the generated file:
>From
609e1d1840da25614a7c9e8954e5356050c9f2ad Mon Sep 17 00:00:00 2001
From: Jim Meyering <meyering@redhat.com>
Date: Wed, 29 Jul 2009 08:13:35 -0400
Subject: [PATCH] build: avoid locale-specific changes in generated, VC'd file
* Makefile.am (all-local): Use LC_ALL=C to sort in C locale.
* po/POTFILES.in: Regenerate.
Richard W.M. Jones [Wed, 29 Jul 2009 15:01:23 +0000 (16:01 +0100)]
On RHEL 5, 'file' command prints 'AMD x86-64' for 'x86-64' arch.
Richard Jones [Wed, 29 Jul 2009 14:55:17 +0000 (15:55 +0100)]
Run cpio with --quiet option so it doesn't print 'xx blocks'
Richard Jones [Wed, 29 Jul 2009 14:49:48 +0000 (15:49 +0100)]
Lib.pm: Skip tests if perl-libintl module is not available.
Richard Jones [Wed, 29 Jul 2009 14:33:22 +0000 (15:33 +0100)]
Lib.pm: Use Perl backtick instead of slurp (Aron Griffis).
Richard Jones [Wed, 29 Jul 2009 14:02:16 +0000 (15:02 +0100)]
Don't show empty CD devices (RHBZ#514505).
Richard Jones [Wed, 29 Jul 2009 11:45:11 +0000 (12:45 +0100)]
ocaml: Rebuild the tests from source if the main library changes.
Richard Jones [Wed, 29 Jul 2009 08:48:08 +0000 (09:48 +0100)]
inspector: Determine guest architecture.
This commit adds kernel and userspace architecture detection
(fully for Linux, partially for Windows). It adds an architecture
for each kernel detected, and an architecture for each root (ie.
userspace) found.
Richard Jones [Tue, 28 Jul 2009 16:55:35 +0000 (17:55 +0100)]
Lib.pm: Add file_architecture command.
This command detects the architecture of some types of binaries,
libraries, kernel modules and initrd images.
Richard Jones [Wed, 29 Jul 2009 09:15:06 +0000 (10:15 +0100)]
Lib: Document the $ro parameter for mount_operating_system properly.
Richard Jones [Wed, 29 Jul 2009 08:39:02 +0000 (09:39 +0100)]
Mention related tools in the manual page.
Richard Jones [Tue, 28 Jul 2009 16:14:14 +0000 (17:14 +0100)]
Whitespace: indent some names in images/Makefile.am correctly.
Richard Jones [Tue, 28 Jul 2009 14:43:14 +0000 (15:43 +0100)]
Lib.pm: Use 'file' as replacement for 'zfile'.
Richard Jones [Tue, 28 Jul 2009 14:29:23 +0000 (15:29 +0100)]
Pass '-z' parameter to 'file' command so it looks inside compressed files.
Also we deprecate the old 'zfile' command.
Richard Jones [Tue, 28 Jul 2009 14:24:44 +0000 (15:24 +0100)]
Add DeprecatedBy flag to mark functions which are deprecated.
Richard Jones [Tue, 28 Jul 2009 14:23:28 +0000 (15:23 +0100)]
Need to declare asprintf_nowarn as inline to avoid gcc complaining.
Richard Jones [Mon, 27 Jul 2009 21:27:45 +0000 (22:27 +0100)]
Replace shell_quote function with %Q and %R printf specifiers.
%Q => simple shell quoted string
%R => path will be prefixed by /sysroot
eg. snprintf (cmd, sizeof cmd, "cat %R", path); system (cmd);
Richard Jones [Mon, 27 Jul 2009 15:39:02 +0000 (16:39 +0100)]
Add InitSquashFS test type, for tests that just use the squashfs (/dev/sdd).
This also speeds up those tests because we don't have to wait
to create a filesystem which is never used.
Matthew Booth [Fri, 24 Jul 2009 13:56:38 +0000 (14:56 +0100)]
Make read-only optional in mount_operating_system()
Richard W.M. Jones [Fri, 24 Jul 2009 17:29:42 +0000 (18:29 +0100)]
More ideas on a quick Perl one-liner syntax for TODO file.
Richard W.M. Jones [Fri, 24 Jul 2009 14:46:15 +0000 (15:46 +0100)]
guestfish -iv should print virt-inspector command (for debugging).
Richard W.M. Jones [Fri, 24 Jul 2009 13:36:50 +0000 (14:36 +0100)]
Reformat the TODO file.
Richard Jones [Fri, 24 Jul 2009 08:31:49 +0000 (09:31 +0100)]
Pass cgroup_disable=memory, saves ~ 5MB of RAM.
Richard Jones [Thu, 23 Jul 2009 17:57:05 +0000 (18:57 +0100)]
get_append call can return NULL, but bindings didn't handle it.
It's not a good idea because there is no way to return an error
indication. However at least this means it won't segfault.
Richard Jones [Thu, 23 Jul 2009 11:45:12 +0000 (12:45 +0100)]
RHEL 5 thinks squashfs is HFS+ filesystem, unless we specify the type explicitly.