Bug#1037541: libguestfs: FTBFS: supermin: error: lstat: Value too large for defined data type: /var/tmp/supermin2b73c8.tmpdir/base.d/init
[https://bugs.debian.org/cgi-bin/bugreport.cgi?bug=1037541] I think what's happening here is we are using the 31/32 bit lstat functions in OCaml: https://github.com/ocaml/ocaml/blob/3d50c168062e9fdbacf739a73738b10108ff3628/otherlibs/unix/unix.mli#L500 (I guess st_size which is signed 31 bits on armv7, but 63 bits on 64 bit platforms). We ought to be using the 64 bit version instead: https://github.com/ocaml/ocaml/blob/3d50c168062e9fdbacf739a73738b10108ff3628/otherlibs/unix/unix.mli#L544 The specific problem will be that the size of some file has grown larger than 1GB resulting in this runtime error, although that seems unusual. Now we *ought* to be using those functions already because we should always be doing: open Unix open Unix.LargeFile to use the "LargeFile" (64 bit) versions of the functions. I had a quick look through the code to see where we are using 'open Unix' without 'open Unix.LargeFile' but I couldn't find the place, leaving me confused. (Of course the other problem is that Fedora has abandoned all 32 bit platforms long ago so we never test this ...) Practically my recommendation is to exclude the whole virt stack on armv7 since it has been years since we tested it. Rich. -- Richard Jones, Virtualization Group, Red Hat http://people.redhat.com/~rjones Read my programming and virtualization blog: http://rwmj.wordpress.com libguestfs lets you edit virtual machines. Supports shell scripting, bindings from many languages. http://libguestfs.org
Bug#1002986: libguestfs-tools: Depends on guestfs-tools that is not in the archive
> It looks like libguestfs-tools version 1:1.46.2-1 in depending on > guestfs-tools that is not in the archive making the package > uninstalable Just wanted to note that upstream we have split up the old, very large, libguestfs git "monorepo", splitting out: guestfs-tools (https://github.com/rwmjones/guestfs-tools) - All the C and OCaml tools like virt-df, virt-customize, etc. except for guestfish, virt-v2v and virt-p2v. virt-v2v (https://github.com/libguestfs/virt-v2v) - Conversion from VMware to KVM. virt-p2v (https://github.com/libguestfs/virt-p2v) - Conversion from physical machines to KVM. libguestfs (https://github.com/libguestfs/libguestfs) - Now it's just the library and language bindings (including guestfish because that is -- kind of -- a set of shell script bindings) FWIW here are the new Fedora packages in case you want to see how we decided to arrange the subpackages and dependencies. https://src.fedoraproject.org/rpms/libguestfs/tree/rawhide https://src.fedoraproject.org/rpms/guestfs-tools/tree/rawhide https://src.fedoraproject.org/rpms/virt-v2v/tree/rawhide https://src.fedoraproject.org/rpms/virt-p2v/tree/rawhide Another thing to say is that we're planning eventually to move all the git repos to gitlab because it's a more free software friendly solution compared to github. Rich. -- Richard Jones, Virtualization Group, Red Hat http://people.redhat.com/~rjones Read my programming and virtualization blog: http://rwmj.wordpress.com virt-top is 'top' for virtual machines. Tiny program with many powerful monitoring features, net stats, disk stats, logging, etc. http://people.redhat.com/~rjones/virt-top
Bug#513451: qemu-img: add stdin/stdout support
12 years after this bug was filed, and because this is the top hit for "qemu-img" and stdin / stdout, I want to point out that this is now possible in Debian and other Linux distros using nbdcopy (in package libnbd-bin) + qemu-nbd. To stream from any qemu format such as qcow2 out to stdout: nbdcopy -- [ qemu-nbd -f qcow2 image.qcow2 ] - | file - To stream from stdin into any qemu format such as qcow2: zcat image.gz | nbdcopy -- - [ qemu-nbd -f qcow2 output.qcow2 ]
Bug#972241: FTBFS on amd64 and i386 (error: guestfs_launch failed)
guestfsd: error while loading shared libraries: libtirpc.so.3: cannot open shared object file: No such file or directory Did something change with how libtirpc gets packaged on Debian or upstream? Rich. -- Richard Jones, Virtualization Group, Red Hat http://people.redhat.com/~rjones Read my programming and virtualization blog: http://rwmj.wordpress.com virt-p2v converts physical machines to virtual machines. Boot with a live CD or over the network (PXE) and turn machines into KVM guests. http://libguestfs.org/virt-v2v
Bug#967099: libguestfs: FTBFS: dwz: debian/guestfsd/usr/sbin/guestfsd: DWARF version 0 unhandled
This is actually a bug in binutils, not OCaml. The fix is: https://sourceware.org/git/?p=binutils-gdb.git;a=commitdiff;h=4d8ee860737005517be588f4771c358593fa421c See also: https://lists.fedoraproject.org/archives/list/de...@lists.fedoraproject.org/message/3YRZ5TJK7PTYDYHUDOYC5HFWKZPA7KIJ/ Rich. -- Richard Jones, Virtualization Group, Red Hat http://people.redhat.com/~rjones Read my programming and virtualization blog: http://rwmj.wordpress.com virt-p2v converts physical machines to virtual machines. Boot with a live CD or over the network (PXE) and turn machines into KVM guests. http://libguestfs.org/virt-v2v
Bug#969009: transition: sleuthkit
The libguestfs FTBFS thing is caused by a bug in binutils (not OCaml or libguestfs). See: https://lists.fedoraproject.org/archives/list/de...@lists.fedoraproject.org/message/3YRZ5TJK7PTYDYHUDOYC5HFWKZPA7KIJ/ https://sourceware.org/git/?p=binutils-gdb.git;a=commitdiff;h=4d8ee860737005517be588f4771c358593fa421c Rich. -- Richard Jones, Virtualization Group, Red Hat http://people.redhat.com/~rjones Read my programming and virtualization blog: http://rwmj.wordpress.com virt-top is 'top' for virtual machines. Tiny program with many powerful monitoring features, net stats, disk stats, logging, etc. http://people.redhat.com/~rjones/virt-top
Bug#946594: Fwd: libguestfs incorrectly detects host CPU architecture
On Thu, Dec 12, 2019 at 11:01:02AM +0100, Vincent Danjean wrote: > [resend to the good (cloned) bug, sorry for the message in the original bug, > it was a mistake] > > Hi, > > On Thu, 5 Dec 2019 11:12:56 +0000 "Richard W.M. Jones" > wrote: > > I believe this is a new bug and nothing much to do with: > > https://bugs.debian.org/cgi-bin/bugreport.cgi?bug=775761 > > So I just cloned the bug into #946594 in order to manage the bug reported > by Pierre Neyron. > > > However about this new bug, what is supposed to happen is that > > common/mlstdutils/guestfs_config.ml is generated by ./configure with > > the correct @host_cpu@ substituted. If that's not happening then it's > > a build issue on Debian of some kind. > > I closely look at the build system with him. It occurs that : > - the Debian package use an 'out-of-source' build (ie the build is > *not* done into the source directories) > - Debian applies some patches to archive this > - but the current rules about guestfs_config.ml especially target > the source directory (whereas ./configure generates it in the > build directory, as for all other configure generated files) > > Easy test : > - remove common/mlstdutils/guestfs_config.ml in the source directory > - build the Debian package (dpkg-buildpackage -us -uc) > => the build fails with: > ocamlfind ocamlopt -g -annot -safe-string -warn-error CDEFLMPSUVYZX+52-3 > -ccopt '-g -O2 > -fdebug-prefix-map=/home/polaris/danjean/libguestfs/libguestfs=. > -fstack-protector-strong -Wformat -Werror=format-security > -fno-strict-overflow -Wno-strict-overflow' -package str,unix -I . -c > ../../../../common/mlstdutils/std_utils.ml -o std_utils.cmx > File "_none_", line 1: > Warning 58: no cmx file was found in path for module Guestfs_config, and its > interface was not compiled with -opaque > > => the guestfs_config.ml generated in the build directory is not > taken into account. > => the Debian package is currently using the provided (x86-64) > guestfs_config.ml in the source directory and ignore the generated > one (in the build directory) on all architectures > => the Debian package is correctly built only on x86-64... > > The root of the problem is in subdir-rules.mk at the root of the > package. I found a fix the seems to work: > = > --- subdir-rules.mk 2019-12-11 15:45:01.274572831 +0100 > +++ subdir-rules.mk.fixed 2019-12-11 15:44:23.738419530 +0100 > @@ -79,12 +79,12 @@ > guestfs_am_v_jar_ = $(guestfs_am_v_jar_@AM_DEFAULT_V@) > guestfs_am_v_jar_0 = @echo " JAR " $@; > > -%.cmi: $(srcdir)/%.mli > +%.cmi: %.mli > $(guestfs_am_v_ocamlcmi)$(OCAMLFIND) ocamlc $(OCAMLFLAGS) > $(OCAMLPACKAGES) -c $< -o $@ > -%.cmo: $(srcdir)/%.ml > +%.cmo: %.ml > $(guestfs_am_v_ocamlc)$(OCAMLFIND) ocamlc $(OCAMLFLAGS) > $(OCAMLPACKAGES) -c $< -o $@ > if HAVE_OCAMLOPT > -%.cmx: $(srcdir)/%.ml > +%.cmx: %.ml > $(guestfs_am_v_ocamlopt)$(OCAMLFIND) ocamlopt $(OCAMLFLAGS) > $(OCAMLPACKAGES) -c $< -o $@ > endif > = > > The idea is to let "make" to check itself in the build directory and then > in the source directory by not forcing a path and using the VPATH feature > (as for all internal automake rules) Pino, Hillu, what do you think? Rich. > Looking at the ChangeLog, I saw that the build rules about cmi/cmo/cmx/... > seems tricky, so I would prefer that someone that know ocaml builds checks > my patch. > In any case, the Debian build is broken. > As upstream does not seem to support out-of-source build, this problem > should > not appear for it. So the fix can go into a debian patch (even if I think that > my patch is a no-op for a 'in-source' build) > > Regards, > Vincent > > > > Rich. > > > > -- > > Richard Jones, Virtualization Group, Red Hat > > http://people.redhat.com/~rjones > > Read my programming and virtualization blog: http://rwmj.wordpress.com > > Fedora Windows cross-compiler. Compile Windows programs, test, and > > build Windows installers. Over 100 libraries supported. > > http://fedoraproject.org/wiki/MinGW > > > > > > > > -- > Vincent Danjean GPG key ID 0xD17897FA vdanj...@debian.org > GPG key fingerprint: 621E 3509 654D D77C 43F5 CA4A F6AE F2AF D178 97FA > Unofficial pkgs: http://moais.imag.fr/membres/vincent.danjean/deb.html > APT repo: deb http://people.debian.org/~vdanjean/debian unstable main > > -- > Vincent Danjean GPG key ID 0xD17897FA vdanj...@debian.org > GPG key fingerprint: 621E 3509 654D D77C 43F5 CA4A F6AE F2AF
Bug#775761: libguestfs incorrectly detects host CPU architecture
I believe this is a new bug and nothing much to do with: https://bugs.debian.org/cgi-bin/bugreport.cgi?bug=775761 However about this new bug, what is supposed to happen is that common/mlstdutils/guestfs_config.ml is generated by ./configure with the correct @host_cpu@ substituted. If that's not happening then it's a build issue on Debian of some kind. Rich. -- Richard Jones, Virtualization Group, Red Hat http://people.redhat.com/~rjones Read my programming and virtualization blog: http://rwmj.wordpress.com Fedora Windows cross-compiler. Compile Windows programs, test, and build Windows installers. Over 100 libraries supported. http://fedoraproject.org/wiki/MinGW
Bug#939546: libnbd: please make the build reproducible
Upstream in: https://github.com/libguestfs/libnbd/commit/2a955696cde1be60d468fdc662b4f70ec862b866 which will be in libnbd >= 1.0.1 & >= 1.1. Rich. -- Richard Jones, Virtualization Group, Red Hat http://people.redhat.com/~rjones Read my programming and virtualization blog: http://rwmj.wordpress.com Fedora Windows cross-compiler. Compile Windows programs, test, and build Windows installers. Over 100 libraries supported. http://fedoraproject.org/wiki/MinGW
Bug#888182: qemu-system-arm: qemu-system-arm crashes when run as part of the nbdkit testsuite
I was able to reproduce a crashing bug in qemu-system-arm on armv7 host. I'm _not_ going to claim this is the same bug that Debian is seeing, but it might be. It's deep inside TCG and unfortunately there is not a lot of useful information in the stack trace. However it's clearly a bug in qemu. Core was generated by `/usr/bin/qemu-system-arm -global virtio-blk-device.scsi=off -enable-fips -nodef'. Program terminated with signal SIGSEGV, Segmentation fault. #0 0xab6a7aa8 in ?? () [Current thread is 1 (Thread 0xaa9f7f90 (LWP 25252))] (gdb) bt #0 0xab6a7aa8 in ?? () #1 0xaba87ad0 in code_gen_buffer () #2 0x006291e0 in cpu_tb_exec (itb=, itb=, cpu=) at /usr/src/debug/qemu-2.11.0-4.fc28.arm/accel/tcg/cpu-exec.c:167 #3 cpu_loop_exec_tb (tb_exit=, last_tb=, tb=, cpu=) at /usr/src/debug/qemu-2.11.0-4.fc28.arm/accel/tcg/cpu-exec.c:627 #4 cpu_exec (cpu=) at /usr/src/debug/qemu-2.11.0-4.fc28.arm/accel/tcg/cpu-exec.c:736 #5 0x005efae4 in qemu_tcg_cpu_thread_fn (arg=0x13ce3e0) at /usr/src/debug/qemu-2.11.0-4.fc28.arm/cpus.c:1270 #6 0xb53f3f1c in start_thread () from /lib/libpthread.so.0 #7 0xb53790d8 in ?? () from /lib/libc.so.6 Backtrace stopped: previous frame identical to this frame (corrupt stack?) I'm using: * qemu-system-arm-2.11.0-4.fc28.armv7hl * kernel-lpae-4.15.0-0.rc9.git2.1.fc28.armv7hl Rich. -- Richard Jones, Virtualization Group, Red Hat http://people.redhat.com/~rjones Read my programming and virtualization blog: http://rwmj.wordpress.com virt-p2v converts physical machines to virtual machines. Boot with a live CD or over the network (PXE) and turn machines into KVM guests. http://libguestfs.org/virt-v2v
Bug#888182: qemu-system-arm: qemu-system-arm crashes when run as part of the nbdkit testsuite
The test failure looks like a problem in libguestfs (or maybe qemu) rather than nbdkit. TBH in Fedora we disable nbdkit tests on 32-bit armv7, 32-bit i686 and all POWER: https://src.fedoraproject.org/rpms/nbdkit/blob/44518f07e0b28a799fa683f1f5ec2ca9c000ac01/f/nbdkit.spec#_419 Now that's not because they shouldn't work eventually, it's because we don't routinely test nbdkit upstream on these architectures. Rich. -- Richard Jones, Virtualization Group, Red Hat http://people.redhat.com/~rjones Read my programming and virtualization blog: http://rwmj.wordpress.com virt-top is 'top' for virtual machines. Tiny program with many powerful monitoring features, net stats, disk stats, logging, etc. http://people.redhat.com/~rjones/virt-top
Bug#887969: nbdkit FTBFS: test failures
On the face of it, it looks like the following patch should fix it: https://github.com/libguestfs/nbdkit/commit/a6b907e5359a139404cee5fd5c7f94eab36a5fde What I don't understand is why that isn't included in your nbdkit (1.1.25) already? In any case the latest nbdkit -- 1.1.27 released a couple of days ago -- does pass all the tests with the latest qemu. The valgrind tests are broken with 1.1.27, and that is fixed by the following 3 commits added after 1.1.27: https://github.com/libguestfs/nbdkit/commit/1e51756907646bc8d2dbcddb2cdb9cea59e5c7bf https://github.com/libguestfs/nbdkit/commit/a3d4a17971bbd84639437a52f5ab1b3fd307c29d https://github.com/libguestfs/nbdkit/commit/b67a2099105a949b9e943b8c8fc8da1c0163c641 Rich. -- Richard Jones, Virtualization Group, Red Hat http://people.redhat.com/~rjones Read my programming and virtualization blog: http://rwmj.wordpress.com virt-df lists disk usage of guests without needing to install any software inside the virtual machine. Supports Linux and Windows. http://people.redhat.com/~rjones/virt-df/
Bug#849597: libguestfs0: Missing multiple dependencies
In Fedora we package up the icoutils dependencies in a separate subpackage to avoid pulling in all of X and Perl when installing the main library: http://pkgs.fedoraproject.org/cgit/rpms/libguestfs.git/tree/libguestfs.spec#n427 Rich. -- Richard Jones, Virtualization Group, Red Hat http://people.redhat.com/~rjones Read my programming and virtualization blog: http://rwmj.wordpress.com virt-p2v converts physical machines to virtual machines. Boot with a live CD or over the network (PXE) and turn machines into KVM guests. http://libguestfs.org/virt-v2v
Bug#838995: No support for "/usr merge"
I posted a patch upstream which should fix or at least work around this: https://www.redhat.com/archives/libguestfs/2016-September/msg00189.html Rich. -- Richard Jones, Virtualization Group, Red Hat http://people.redhat.com/~rjones Read my programming and virtualization blog: http://rwmj.wordpress.com virt-top is 'top' for virtual machines. Tiny program with many powerful monitoring features, net stats, disk stats, logging, etc. http://people.redhat.com/~rjones/virt-top
Bug#815409: libguestfs: FTBFS on mips: segfault while creating blank-disk-1s.qcow2
I actually have libguestfs running on mipsel at home. It's very slow indeed :-( Anyway, it looks as if the segfault is happening in the `qemu-img' utility. The command which fails is: qemu-img create -f qcow2 -o preallocation=metadata blank-disk-1s.qcow2 512 (You could just run that command on its own on the same hardware to see if you can reproduce it). This bug should probably be reassigned to qemu. Rich. -- Richard Jones, Virtualization Group, Red Hat http://people.redhat.com/~rjones Read my programming and virtualization blog: http://rwmj.wordpress.com libguestfs lets you edit virtual machines. Supports shell scripting, bindings from many languages. http://libguestfs.org
Bug#809185: Note regarding bin2s.pl
On Tue, Jan 12, 2016 at 07:57:03AM +0100, Hilko Bengen wrote: > Helge, > > I have applied all the architecture-specific bits but not the bin2s > script yet. TBH, so far I don't see what is wrong about export and use > of the "_binary_init_size" constant. [https://bugs.debian.org/cgi-bin/bugreport.cgi?bug=809185] I see it as a reasonable simplification - it allows us to get rid of that conditional code for HP-UX in bin2s.pl. However looking at the patch, I don't like the casts in: - size_t n = (size_t) &_binary_init_size; + size_t n = ((size_t) &_binary_init_end) - ((size_t) &_binary_init_start); Since those are pointers, it seems better to simply subtract them. (Though it would be better if we'd declared the type of _binary_init_start/_end as uint8_t instead of char.) If we must cast them then the correct integer to use is 'intptr_t', an int type that's guaranteed by C99 to be long enough to store a pointer. Rich. -- Richard Jones, Virtualization Group, Red Hat http://people.redhat.com/~rjones Read my programming and virtualization blog: http://rwmj.wordpress.com virt-top is 'top' for virtual machines. Tiny program with many powerful monitoring features, net stats, disk stats, logging, etc. http://people.redhat.com/~rjones/virt-top
Bug#809185: [Libguestfs] Note regarding bin2s.pl
On Tue, Jan 12, 2016 at 10:05:00AM +, Richard W.M. Jones wrote:
> On Tue, Jan 12, 2016 at 07:57:03AM +0100, Hilko Bengen wrote:
> > Helge,
> >
> > I have applied all the architecture-specific bits but not the bin2s
> > script yet. TBH, so far I don't see what is wrong about export and use
> > of the "_binary_init_size" constant.
>
> [https://bugs.debian.org/cgi-bin/bugreport.cgi?bug=809185]
>
> I see it as a reasonable simplification - it allows us to get rid of
> that conditional code for HP-UX in bin2s.pl.
>
> However looking at the patch, I don't like the casts in:
>
> - size_t n = (size_t) &_binary_init_size;
> + size_t n = ((size_t) &_binary_init_end) - ((size_t) &_binary_init_start);
>
> Since those are pointers, it seems better to simply subtract them.
> (Though it would be better if we'd declared the type of
> _binary_init_start/_end as uint8_t instead of char.)
>
> If we must cast them then the correct integer to use is 'intptr_t', an
> int type that's guaranteed by C99 to be long enough to store a
> pointer.
How about the attached patch?
Rich.
--
Richard Jones, Virtualization Group, Red Hat http://people.redhat.com/~rjones
Read my programming and virtualization blog: http://rwmj.wordpress.com
virt-df lists disk usage of guests without needing to install any
software inside the virtual machine. Supports Linux and Windows.
http://people.redhat.com/~rjones/virt-df/
>From 7cff794d82076df70dde7a851937fc7bf93fdf44 Mon Sep 17 00:00:00 2001
From: "Richard W.M. Jones" <rjo...@redhat.com>
Date: Tue, 12 Jan 2016 11:07:57 +
Subject: [PATCH] bin2s: Remove _size, since it can be computed from _start and
_end.
Also declare the _start and _end as uint8_t instead of char, since
they refer to an array of bytes.
See discussion here:
https://bugs.debian.org/cgi-bin/bugreport.cgi?bug=809185
---
src/bin2s.pl | 8
src/ext2init-c.c | 5 +++--
2 files changed, 3 insertions(+), 10 deletions(-)
diff --git a/src/bin2s.pl b/src/bin2s.pl
index db6db26..8558126 100755
--- a/src/bin2s.pl
+++ b/src/bin2s.pl
@@ -39,7 +39,6 @@ print $ofh <<"EOF";
\t.globl\t_binary_${infile}_start
\t.globl\t_binary_${infile}_end
-\t.globl\t_binary_${infile}_size
\t.section\t.rodata
_binary_${infile}_start:
@@ -55,14 +54,7 @@ die "read $infile (at offset $sz): $!\n" if not defined $i;
close $ifh;
print $ofh <<"EOF";
-
_binary_${infile}_end:
-
-#if defined(__hppa__)
-\t_binary_${infile}_size: .equ $sz
-#else
-\t.equ _binary_${infile}_size, $sz
-#endif
EOF
close $ofh;
diff --git a/src/ext2init-c.c b/src/ext2init-c.c
index c310ed2..66ad254 100644
--- a/src/ext2init-c.c
+++ b/src/ext2init-c.c
@@ -20,6 +20,7 @@
#include
#include
+#include
#include
#include
@@ -28,14 +29,14 @@
/* The init binary.
* See: bin2s.pl, init.c.
*/
-extern char _binary_init_start, _binary_init_end, _binary_init_size;
+extern uint8_t _binary_init_start, _binary_init_end;
value
supermin_binary_init (value unitv)
{
CAMLparam1 (unitv);
CAMLlocal1 (sv);
- size_t n = (size_t) &_binary_init_size;
+ size_t n = &_binary_init_end - &_binary_init_start;
sv = caml_alloc_string (n);
memcpy (String_val (sv), &_binary_init_start, n);
--
2.5.0
Bug#809185: [Libguestfs] Note regarding bin2s.pl
On Tue, Jan 12, 2016 at 06:21:14PM +0100, Helge Deller wrote: > On 12.01.2016 12:10, Richard W.M. Jones wrote: > > On Tue, Jan 12, 2016 at 10:05:00AM +, Richard W.M. Jones wrote: > >> On Tue, Jan 12, 2016 at 07:57:03AM +0100, Hilko Bengen wrote: > >>> Helge, > >>> > >>> I have applied all the architecture-specific bits but not the bin2s > >>> script yet. TBH, so far I don't see what is wrong about export and use > >>> of the "_binary_init_size" constant. > >> > >> [https://bugs.debian.org/cgi-bin/bugreport.cgi?bug=809185] > >> > >> I see it as a reasonable simplification - it allows us to get rid of > >> that conditional code for HP-UX in bin2s.pl. > >> > >> However looking at the patch, I don't like the casts in: > >> > >> - size_t n = (size_t) &_binary_init_size; > >> + size_t n = ((size_t) &_binary_init_end) - ((size_t) > >> &_binary_init_start); > >> > >> Since those are pointers, it seems better to simply subtract them. > >> (Though it would be better if we'd declared the type of > >> _binary_init_start/_end as uint8_t instead of char.) > >> > >> If we must cast them then the correct integer to use is 'intptr_t', an > >> int type that's guaranteed by C99 to be long enough to store a > >> pointer. > > > > How about the attached patch? > > In general I'd say it looks OK. > Just a few comments: > > -extern char _binary_init_start, _binary_init_end, _binary_init_size; > +extern uint8_t _binary_init_start, _binary_init_end; > > Does the char to uint8_t change really makes such a big difference? > We will just use the address of the variable anyway. It seems from this answer: https://stackoverflow.com/questions/2215445/are-there-machines-where-sizeofchar-1-or-at-least-char-bit-8 that sizeof (char) is always 1 so it would never make a difference. However I suppose it's better to be clearer about what we intend. > - size_t n = (size_t) &_binary_init_size; > + size_t n = &_binary_init_end - &_binary_init_start; > > It's OK, but maybe some compilers/platforms might complain with a warning. > It might be better to keep a cast to (size_t), e.g.: > + size_t n = (size_t) (&_binary_init_end - &_binary_init_start); I tested the patch today on 32- and 64-bit Linux x86 systems, but nothing beyond that. Actually I think if it warns, I want to know about that -- eg. in some bizarre case where size_t isn't sufficient to store the result. Patch applied anyway, thanks. Rich. > But either way, I'm fine with both approaches. > > Thanks! > Helge -- Richard Jones, Virtualization Group, Red Hat http://people.redhat.com/~rjones Read my programming and virtualization blog: http://rwmj.wordpress.com virt-df lists disk usage of guests without needing to install any software inside the virtual machine. Supports Linux and Windows. http://people.redhat.com/~rjones/virt-df/
Bug#796008: /usr/lib/x86_64-linux-gnu/virt-p2v: /usr/lib/x86_64-linux-gnu/virt-p2v installed in incorrect path
Just a note that the upstream p2v/Makefile.am uses: libexec_PROGRAMS = virt-p2v In Fedora, the binary is placed in /usr/libexec (which doesn't exist on Debian of course). Wherever you put it, the command is not intended to be run by any user (neither non-root, nor root). The command should only be used when packaged into a bootable ISO, and in that case it runs from a wrapper script at boot time. Rich.
Bug#794732: libguestfs: FTBFS
It seems to be the same as this Ubuntu bug report: https://bugs.launchpad.net/ubuntu/+source/ocaml-gettext/+bug/1481994 Rich. -- To UNSUBSCRIBE, email to debian-bugs-dist-requ...@lists.debian.org with a subject of unsubscribe. Trouble? Contact listmas...@lists.debian.org
Bug#775514: No networking running virt-builder scripts after upgrading from 1.26 to 1.28
On Tue, Jun 30, 2015 at 09:46:31AM +0200, Hilko Bengen wrote: (2) The second problem is that we run 'dhclient', but for some reason this command cannot acquire an IP address unless you name a specific interface (eg 'dhclient eth0'). This was already reported for Ubuntu: https://bugzilla.redhat.com/show_bug.cgi?id=1224795 The workaround is just this patch (to appliance/init): - dhclient + dhclient eth0 Can we assume that the network name is always eth0? In that case, I'd simply apply that patch. Correct -- that is the problem that stops me from just adding that patch upstream. I guess that different kernel versions, and particularly, different versions of udev rules, could cause that interface to have a different name. `dhclient eth0' seems to work for Debian and Ubuntu. Even if it doesn't work, it won't make things worse than they are now. For upstream, I intend to investigate why `dhclient' alone doesn't work (it works fine on Fedora). Rich. -- Richard Jones, Virtualization Group, Red Hat http://people.redhat.com/~rjones Read my programming and virtualization blog: http://rwmj.wordpress.com Fedora Windows cross-compiler. Compile Windows programs, test, and build Windows installers. Over 100 libraries supported. http://fedoraproject.org/wiki/MinGW -- To UNSUBSCRIBE, email to debian-bugs-dist-requ...@lists.debian.org with a subject of unsubscribe. Trouble? Contact listmas...@lists.debian.org
Bug#775514: No networking running virt-builder scripts after upgrading from 1.26 to 1.28
I went through this with Adam and there are two bugs: (1) isc-dhcp-client is not listed in /usr/lib/x86_64-linux-gnu/guestfs/supermin.d/packages As a result, dhclient is not loaded into the appliance, and the dhclient command fails. I thought this was strange, because upstream we have 'isc-dhcp-client' listed in packagelist.in, but then I realized that the problem is really that 'isc-dhcp-client' is a missing build dependency of libguestfs on Debian. It also needs to be a runtime dependency. Because it's not a build-time dependency, it doesn't get pulled in to the 'packages' file. It also has to be a runtime dependency so that users will have it installed on their machines. (2) The second problem is that we run 'dhclient', but for some reason this command cannot acquire an IP address unless you name a specific interface (eg 'dhclient eth0'). This was already reported for Ubuntu: https://bugzilla.redhat.com/show_bug.cgi?id=1224795 The workaround is just this patch (to appliance/init): - dhclient + dhclient eth0 but I don't know why that works. I believe this to be a bug in dhclient, not in libguestfs. Rich. -- Richard Jones, Virtualization Group, Red Hat http://people.redhat.com/~rjones Read my programming and virtualization blog: http://rwmj.wordpress.com virt-top is 'top' for virtual machines. Tiny program with many powerful monitoring features, net stats, disk stats, logging, etc. http://people.redhat.com/~rjones/virt-top -- To UNSUBSCRIBE, email to debian-bugs-dist-requ...@lists.debian.org with a subject of unsubscribe. Trouble? Contact listmas...@lists.debian.org
Bug#775514: No networking running virt-builder scripts after upgrading from 1.26 to 1.28
Could someone attach the *complete* and *unedited* output (in /tmp/log) of the following commands to the bug: guestfish --network -a /dev/null run -v -x | tee /tmp/log and: virt-builder debian-7 -v -x \ --run-command 'wget -O /dev/null https://www.debian.org/' \ | tee /tmp/log Rich. -- Richard Jones, Virtualization Group, Red Hat http://people.redhat.com/~rjones Read my programming and virtualization blog: http://rwmj.wordpress.com virt-p2v converts physical machines to virtual machines. Boot with a live CD or over the network (PXE) and turn machines into KVM guests. http://libguestfs.org/virt-v2v -- To UNSUBSCRIBE, email to debian-bugs-dist-requ...@lists.debian.org with a subject of unsubscribe. Trouble? Contact listmas...@lists.debian.org
Bug#783038: libguestfs-tools: virt-filesystems output is incorrect
The output of virt-filesystems isn't wrong. virt-resize can't resize extended partitions (like /dev/sda5). See the section LOGICAL PARTITIONS in the virt-resize man page. Rich. -- Richard Jones, Virtualization Group, Red Hat http://people.redhat.com/~rjones Read my programming and virtualization blog: http://rwmj.wordpress.com virt-df lists disk usage of guests without needing to install any software inside the virtual machine. Supports Linux and Windows. http://people.redhat.com/~rjones/virt-df/ -- To UNSUBSCRIBE, email to debian-bugs-dist-requ...@lists.debian.org with a subject of unsubscribe. Trouble? Contact listmas...@lists.debian.org
Bug#694025: Debian package of 'Oz' uploaded to mentors.debian.net
On Fri, Apr 17, 2015 at 01:19:53PM +0200, Simon Josefsson wrote: Do you see any problem including Oz in Debian? The tools look complementary, and Oz provide some value to me. I've added a pointer to the virt-install and virt-builder tools to the Oz package description now. No problem at all - in Fedora we ship both tools. Rich. -- Richard Jones, Virtualization Group, Red Hat http://people.redhat.com/~rjones Read my programming and virtualization blog: http://rwmj.wordpress.com libguestfs lets you edit virtual machines. Supports shell scripting, bindings from many languages. http://libguestfs.org -- To UNSUBSCRIBE, email to debian-bugs-dist-requ...@lists.debian.org with a subject of unsubscribe. Trouble? Contact listmas...@lists.debian.org
Bug#694025: Debian package of 'Oz' uploaded to mentors.debian.net
On Thu, Apr 16, 2015 at 04:35:14PM +0200, Simon Josefsson wrote: Hi. I recently found myself setting up a libvirtd/KVM-based virtual machine, and needed a way to build VM images from the command line. I searched around, and found the Oz project: https://github.com/clalancette/oz/wiki Testing Oz on a newly installed Jessie rc2 machine I found that it seemed to work. At least I was able to create VM images for Debian Wheezy and Jessie easily. Thus I decided to respin the old ITP for Oz: https://bugs.debian.org/cgi-bin/bugreport.cgi?bug=694025 I have uploaded a package to mentors for easy review: https://mentors.debian.net/package/oz If you want to check out the Debian packaging you can get: https://github.com/jas4711/oz-dpkg With this, I'm asking for review/support/objections before uploading this into Debian properly. On sufficiently new Debian you can also do: $ virt-builder debian-7 (and no need for root privileges). Rich. -- Richard Jones, Virtualization Group, Red Hat http://people.redhat.com/~rjones Read my programming and virtualization blog: http://rwmj.wordpress.com libguestfs lets you edit virtual machines. Supports shell scripting, bindings from many languages. http://libguestfs.org -- To UNSUBSCRIBE, email to debian-bugs-dist-requ...@lists.debian.org with a subject of unsubscribe. Trouble? Contact listmas...@lists.debian.org
Bug#694025: Debian package of 'Oz' uploaded to mentors.debian.net
On Thu, Apr 16, 2015 at 07:24:26PM +0200, Simon Josefsson wrote: Thanks for the pointer. Trying it out, it looks to me that they don't offer the same functionality. virt-builder downloads an image that someone else prepared (not sure how?) At the moment using: https://github.com/libguestfs/libguestfs/blob/master/builder/website/debian.sh but I intend to get out of the business of building disk images for virt-builder, and use the cloud images prepared by distros. and adapts it. oz-install creates a new image by booting a normal d-i ISO together with preseeding. Yup, that is basically correct. Of course virt-builder is way faster. Is it possible to do that with virt-builder? No -- although you could run the script above, building your own images for your own virt-builder template repository, and then use virt-builder to quickly deploy them. For my purposes, oz-install's approach is even better than virt-install's approach of doing debootstrapping, since I want the full D-I experience. So these tools look complementary to me. Correct. To reduce confusion, maybe a small discussion in the package description or README.Debian is appropriate to explain the differences. Btw, thanks for doing the Oz Debian packaging. Was there any reason it wasn't uploaded back then? That's a good question -- I don't actually recall at all :-/ Rich. -- Richard Jones, Virtualization Group, Red Hat http://people.redhat.com/~rjones Read my programming and virtualization blog: http://rwmj.wordpress.com Fedora Windows cross-compiler. Compile Windows programs, test, and build Windows installers. Over 100 libraries supported. http://fedoraproject.org/wiki/MinGW -- To UNSUBSCRIBE, email to debian-bugs-dist-requ...@lists.debian.org with a subject of unsubscribe. Trouble? Contact listmas...@lists.debian.org
Bug#777548: libguestfs-tools: virt-builder fails with --size option
The commit that you point to has been backported to 1.28.6 (ie. along the 1.28 stable branch), so updating should be sufficient to fix this. I tested with virt-builder 1.28.6 and the bug has been fixed. Rich. -- Richard Jones, Virtualization Group, Red Hat http://people.redhat.com/~rjones Read my programming and virtualization blog: http://rwmj.wordpress.com virt-top is 'top' for virtual machines. Tiny program with many powerful monitoring features, net stats, disk stats, logging, etc. http://people.redhat.com/~rjones/virt-top -- To UNSUBSCRIBE, email to debian-bugs-dist-requ...@lists.debian.org with a subject of unsubscribe. Trouble? Contact listmas...@lists.debian.org
Bug#775761: libguestfs incorrectly detects host CPU architecture
I wonder why it chose to run /usr/bin/qemu-system-i386 ? Is $LIBGUESTFS_HV or $LIBGUESTFS_QEMU set? What is the output of: guestfish get-hv What happens if you do: LIBGUESTFS_HV=/usr/bin/qemu-system-x86_64 libguestfs-test-tool It looks as if libguestfs is configured wrongly, or maybe you've got the 32 bit version of libguestfs installed on your system. Rich. -- Richard Jones, Virtualization Group, Red Hat http://people.redhat.com/~rjones Read my programming and virtualization blog: http://rwmj.wordpress.com virt-df lists disk usage of guests without needing to install any software inside the virtual machine. Supports Linux and Windows. http://people.redhat.com/~rjones/virt-df/ -- To UNSUBSCRIBE, email to debian-bugs-dist-requ...@lists.debian.org with a subject of unsubscribe. Trouble? Contact listmas...@lists.debian.org
Bug#775761: libguestfs incorrectly detects host CPU architecture
On Tue, Jan 20, 2015 at 07:42:10AM -0800, Rogier wrote: It looks as if libguestfs is configured wrongly, or maybe you've got the 32 bit version of libguestfs installed on your system. I do have a 32-bit version (my entire userland is still 32-bit - it used to be the better choice, and as migrating to 64-bit is a major operation, it is not a high priority...). So the 32-bit version of libguestfs assumes that the qemu version of choice must also be 32 bit, but then it chooses to boot a 64-bit kernel ? You can choose which kernel gets booted by setting up some environment variables. See the manual page here: http://libguestfs.org/supermin.1.html#USING-A-CUSTOM-KERNEL-AND-KERNEL-MODULES Basically just: export SUPERMIN_KERNEL=/boot/my-vmlinuz export SUPERMIN_KERNEL_VERSION=3.XX.0 export SUPERMIN_MODULES=/lib/modules/3.XX.0 Of course, it means you must have a 32 bit kernel for this to work, and supermin will need to find 32 bit binaries and so on. Alternative is to set LIBGUESTFS_HV=/usr/bin/qemu-system-x86_64 which seemed to work for you. Rich. -- Richard Jones, Virtualization Group, Red Hat http://people.redhat.com/~rjones Read my programming and virtualization blog: http://rwmj.wordpress.com virt-df lists disk usage of guests without needing to install any software inside the virtual machine. Supports Linux and Windows. http://people.redhat.com/~rjones/virt-df/ -- To UNSUBSCRIBE, email to debian-bugs-dist-requ...@lists.debian.org with a subject of unsubscribe. Trouble? Contact listmas...@lists.debian.org
Bug#763741: libguestfs: Incorrect build-dependencies (libsystemd-id128-0, libsystemd-journal0), fails binNMU
On Tue, Oct 07, 2014 at 11:44:00PM +0200, Andreas Henriksson wrote: Hello Rich. On Sun, Oct 05, 2014 at 01:59:46PM +0100, Richard W.M. Jones wrote: libaugeas0, libcap2, libhivex0, libpcre3, libxml2, libyajl2, Why are these wrong? They all seem to be correct build deps to me. A debian package normally build-depends on the development package providing the development files (headers, pkg-config, (unversioned) shared object file, etc.) rather then simply on a package with only a particular versioned .so.X. In other words, build-dependencies should normally end with -dev (if it's a library and not a tool used during the build process). The above list would be more correct as: libaugeas-dev, libcap-dev, libhivex-dev, libpcre3-dev, libxml2-dev, libyajl-dev, All of these are probably pulled in indirectly via some other build-dependency right now but should be explicitly specified if they really are build-dependencies of libguestfs. The above -dev packages also specifies a dependency on their respective libfooX package as needed. Doing this correctly is what makes it possible to do binNMUs when there's a transition for a certain library (ie. it gets a new so version). Doing this incorrectly makes that impossible. But in the case of libguestfs, on Debian we depend on specific SONAMEs (not for any real reason, just because Debian's packaging system forces us to do that). So if an SONAME changes it requires an upstream patch or a downstream patch that eventually has to go upstream. So I still think that Hilko's build dependencies are correct. Rich. -- Richard Jones, Virtualization Group, Red Hat http://people.redhat.com/~rjones Read my programming and virtualization blog: http://rwmj.wordpress.com Fedora Windows cross-compiler. Compile Windows programs, test, and build Windows installers. Over 100 libraries supported. http://fedoraproject.org/wiki/MinGW -- To UNSUBSCRIBE, email to debian-bugs-dist-requ...@lists.debian.org with a subject of unsubscribe. Trouble? Contact listmas...@lists.debian.org
Bug#763741: libguestfs: Incorrect build-dependencies (libsystemd-id128-0, libsystemd-journal0), fails binNMU
libaugeas0, libcap2, libhivex0, libpcre3, libxml2, libyajl2, Why are these wrong? They all seem to be correct build deps to me. Rich. -- Richard Jones, Virtualization Group, Red Hat http://people.redhat.com/~rjones Read my programming and virtualization blog: http://rwmj.wordpress.com virt-df lists disk usage of guests without needing to install any software inside the virtual machine. Supports Linux and Windows. http://people.redhat.com/~rjones/virt-df/ -- To UNSUBSCRIBE, email to debian-bugs-dist-requ...@lists.debian.org with a subject of unsubscribe. Trouble? Contact listmas...@lists.debian.org
Bug#763741: libguestfs: Incorrect build-dependencies (libsystemd-id128-0, libsystemd-journal0), fails binNMU
I'm not complete clear on what change is being made to Debian. However from the build log, the problem is that 'libsystemd.so.0' isn't making it into the appliance. This seems to be because the following upstream commit shoud be backported to the 1.26 branch (either in Debian or upstream): commit a36dde98899ea160a49c9af65a883383496cd331 Author: Hilko Bengen ben...@hilluzination.de Date: Fri Sep 19 18:55:50 2014 +0200 appliance: Add libsystemd0 (and systemd as an alternative for sysvinit) on Debian-based systems Rich. -- Richard Jones, Virtualization Group, Red Hat http://people.redhat.com/~rjones Read my programming and virtualization blog: http://rwmj.wordpress.com virt-p2v converts physical machines to virtual machines. Boot with a live CD or over the network (PXE) and turn machines into KVM guests. http://libguestfs.org/virt-v2v -- To UNSUBSCRIBE, email to debian-bugs-dist-requ...@lists.debian.org with a subject of unsubscribe. Trouble? Contact listmas...@lists.debian.org
Bug#754614: libguestfs: FTBFS on i386: Value too large for defined data type
This is a bug in mke2fs, not in libguestfs. See: https://bugzilla.redhat.com/show_bug.cgi?id=1099892 Rich. -- Richard Jones, Virtualization Group, Red Hat http://people.redhat.com/~rjones Read my programming and virtualization blog: http://rwmj.wordpress.com libguestfs lets you edit virtual machines. Supports shell scripting, bindings from many languages. http://libguestfs.org -- To UNSUBSCRIBE, email to debian-bugs-dist-requ...@lists.debian.org with a subject of unsubscribe. Trouble? Contact listmas...@lists.debian.org
Bug#754615: libguestfs: FTBFS on mips: make[2]: *** [quickcheck] Alarm clock
I'm slightly surprised that libguestfs ever worked on MIPS. We've certainly never tested it upstream. The error is a hang in the appliance, which eventually causes libguestfs-test-tool to hit its default timeout (600 seconds). It's not really possible to tell what is causing the hang. We'd need to have access to MIPS hardware to debug this further. However it happens just after starting udev, which may or may not be related. Here is the code: https://github.com/libguestfs/libguestfs/blob/master/appliance/init#L50 Rich. -- Richard Jones, Virtualization Group, Red Hat http://people.redhat.com/~rjones Read my programming and virtualization blog: http://rwmj.wordpress.com libguestfs lets you edit virtual machines. Supports shell scripting, bindings from many languages. http://libguestfs.org -- To UNSUBSCRIBE, email to debian-bugs-dist-requ...@lists.debian.org with a subject of unsubscribe. Trouble? Contact listmas...@lists.debian.org
Bug#710579: libguestfs0: possible to allow alternatives to vim-tiny ?
FYI, on Fedora we have a 'libguestfs-rescue' sub-package which pulls in some extra dependencies that are useful in the virt-rescue shell. This allows us to have larger dependencies for people who want to use virt-rescue, without needing those dependencies elsewhere. I'm not sure if this solves the problem, but that's what we do anyhow. Reference: http://pkgs.fedoraproject.org/cgit/libguestfs.git/tree/libguestfs.spec?id=8249a1e58610ac33bedfd301aad6112b2b8d2f05#n305 Rich. -- To UNSUBSCRIBE, email to debian-bugs-dist-requ...@lists.debian.org with a subject of unsubscribe. Trouble? Contact listmas...@lists.debian.org
Bug#742283: supermin: internal error: assertion failed at dpkg.ml, line 86, char 19
I am able to reproduce this bug with supermin 5.1.5 by doing:
$ sudo dpkg --add-architecture i386
$ sudo apt-get update
$ sudo apt-get install attr:i386
$ sudo apt-get install xfsprogs
(Note that xfsprogs requires attr.)
After that:
$ dpkg-query --show --showformat '${Package} ${Version} ${Architecture}\n' |
grep attr
attr 1:2.4.47-1 i386
libattr1 1:2.4.47-1 amd64
libattr1 1:2.4.47-1 i386
libattr1-dev 1:2.4.47-1 amd64
Now to reproduce the bug, do:
$ supermin --version
supermin 5.1.5
$ supermin --prepare attr xfsprogs -o /tmp/supermin.d
supermin: internal error: assertion failed at dpkg.ml, line 86, char 19
However, using supermin 5.1.6 does not produce the assertion failure,
and I am able to prepare and build an appliance:
$ git describe
v5.1.6
$ ./src/supermin --prepare attr xfsprogs -o /tmp/supermin.d
$ ./src/supermin --build /tmp/supermin.d -o /tmp/appliance.d -f chroot
So I think you should try upgrading to 5.1.6 to fix this bug.
Rich.
--
Richard Jones, Virtualization Group, Red Hat http://people.redhat.com/~rjones
Fedora Windows cross-compiler. Compile Windows programs, test, and
build Windows installers. Over 100 libraries supported.
http://fedoraproject.org/wiki/MinGW
--
To UNSUBSCRIBE, email to debian-bugs-dist-requ...@lists.debian.org
with a subject of unsubscribe. Trouble? Contact listmas...@lists.debian.org
Bug#681889: libguestfs0 has too many dependencies?
Since I wrote the message #26 above, we've done a lot of work upstream and -- starting with libguestfs = 1.26 and supermin = 5 -- it will be possible to split the dependencies of the appliance. The basic ideas are covered in these blog posts: http://rwmj.wordpress.com/2014/02/26/new-in-libguestfs-1-25-38/#content http://rwmj.wordpress.com/2014/03/08/supermin-version-5/#content http://rwmj.wordpress.com/2014/03/20/analysis-of-the-size-of-libguestfs-dependencies/#content The Fedora Rawhide package has already been split into: - libguestfs-gfs2 - libguestfs-hfsplus - libguestfs-jfs - libguestfs-nilfs - libguestfs-reiserfs - libguestfs-rescue # vim dependency isolated here - libguestfs-rsync - libguestfs-xfs - libguestfs-zfs # zfs dependency isolated here The idea is that you install libguestfs base package. If you need to handle guests / disk images using (eg) XFS, then you have to install libguestfs-xfs as well. The base package doesn't pull in some of the larger / troublesome dependencies, so you only need to install those if you need the feature. I also posted a patch to the Debian package which begins implementing this split (only ZFS so far, but the principle is the same for all subpackages): https://www.redhat.com/archives/libguestfs/2014-March/msg00171.html Note this does require the new upstream versions of libguestfs supermin. libguestfs 1.26 should be released at the end of this month, but there are 1.25.x packages in Debian experimental. Rich. -- Richard Jones, Virtualization Group, Red Hat http://people.redhat.com/~rjones Read my programming blog: http://rwmj.wordpress.com Fedora now supports 80 OCaml packages (the OPEN alternative to F#) -- To UNSUBSCRIBE, email to debian-bugs-dist-requ...@lists.debian.org with a subject of unsubscribe. Trouble? Contact listmas...@lists.debian.org
Bug#640328: kvm: Please make /dev/kvm world-accessible in 60-qemu-kvm.rules
In Fedora we have had /dev/kvm mode 0666 for years. It was changed that way in July 2009. There has never been a security problem attributable to this. There is no problem with having any user create a VM. Virtual machines are regular processes, subject to all the usual ulimits. They use regular malloc'd memory. Furthermore small appliances are useful for security: for sandboxing single tasks, and for securely accessing filesystems. Rich. -- To UNSUBSCRIBE, email to debian-bugs-dist-requ...@lists.debian.org with a subject of unsubscribe. Trouble? Contact listmas...@lists.debian.org
Bug#640328: kvm: Please make /dev/kvm world-accessible in 60-qemu-kvm.rules
On Thu, Oct 10, 2013 at 10:18:53PM +0400, Michael Tokarev wrote: 10.10.2013 17:44, Richard W.M. Jones wrote: In Fedora we have had /dev/kvm mode 0666 for years. It was changed that way in July 2009. There has never been a security problem attributable to this. This is plain wrong. Here's a very recent example: http://www.openwall.com/lists/oss-security/2013/08/26/3 This causes [on ARM only] the program running the bad ioctl to oops. (I see no evidence of this causing host OS crash -- just an oops in the calling process). Obviously you fix kernel bugs, and you would have to fix this one anyway. Now compare this to the large number of filesystem bugs around, which can be prevented by only mounting filesystems in appliances. Rich. -- To UNSUBSCRIBE, email to debian-bugs-dist-requ...@lists.debian.org with a subject of unsubscribe. Trouble? Contact listmas...@lists.debian.org
Bug#681889: libguestfs0 has too many dependencies?
To: Thomas Koch: I was forced to install vim! This is insulting for an emacs user! guestfish needs vim to implement the 'vi' command: http://libguestfs.org/guestfish.1.html#vi (and emacs to implement the 'emacs' command). I don't know why libguestfs0 would depend on this however. Seriously. The dependency list of libguestfs0 is very surprising. Could you please check whether libguestfs0 could be split up in smaller binary packages with only minimal dependencies? If that shouldn't be possible (how?), could you please add a short explanation of the dependencies in the package description and a longer one in a README.Debian? This is a pretty common question, and therefore I have added the following section to the FAQ http://libguestfs.org/guestfs-faq.1.html#libguestfs-has-a-really-long-list-of-dependencies- section: Libguestfs has a really long list of dependencies It would be possible to package libguestfs for Debian differently, and although I doubt that you'd be happier with the results, let me know if any of the alternatives suggested in that section are better for you. One thing which is unfortunately *not* possible at this time is to split libguestfs up into different capabilities, eg. enabling only XFS or NTFS by installing a particular subpackage. Having said that, in the Debian package right now there are probably dependencies which are not needed. In particular I don't think that the appliance dependencies are needed on the main package. To: Laurent Bigonville: I might be wrong here, but isn't libguestfs running all the commands into an appliance? Shouldn't all these dependencies installed only inside the appliance? Or is there a mode where libguestfs can run commands on the host? Not quite either of these. Libguestfs builds an appliance using files from the host (eg. /sbin/mkfs), then runs these commands in the appliance only, never on the host. It therefore needs them to be present on the host, so it can copy them into the appliance to run them. The supermin man page explains all this, see: http://libguestfs.org/supermin.8.html#SUPERMIN-APPLIANCES (Note that supermin was previously called febootstrap; it's the same thing) - - - I hope this at least explains some of what's going on, even if it doesn't help much :-( Rich. -- Richard Jones, Virtualization Group, Red Hat http://people.redhat.com/~rjones virt-df lists disk usage of guests without needing to install any software inside the virtual machine. Supports Linux and Windows. http://people.redhat.com/~rjones/virt-df/ -- To UNSUBSCRIBE, email to debian-bugs-dist-requ...@lists.debian.org with a subject of unsubscribe. Trouble? Contact listmas...@lists.debian.org
Bug#699703: libguestfs0: make febootstrap a recommends
There seems to be some confusion and misunderstanding in this bug report. First the facts: (1) febootstrap is a general tool to build supermin appliances. It has nothing to do with 'Fedora' -- the name is just an unfortunate historical accident. (2) On Fedora we split out 'febootstrap' (the part for building supermin appliances from repositories) and 'febootstrap-supermin-helper' (the part for reconstructing supermin appliances at runtime) as separate packages. (3) On Fedora, libguestfs only depends at runtime on febootstrap-supermin-helper which is a small C program that takes next to no space. febootstrap is a build-time requirement only. (4) If you are using supermin appliances, then febootstrap-supermin-helper is very much required, even on Debian. (5) If you are using fixed appliances, then febootstrap-supermin-helper is not involved. https://github.com/libguestfs/libguestfs/blob/master/src/appliance.c#L66 Second some observations arising: (a) AFAIK Debian is still using the supermin appliance by default, although it does build it in a slightly different way from how we do it in Fedora. Nevertheless, febootstrap-supermin-helper is required by Debian (see points (4), (5)). (b) Because Debian builds the supermin appliance from the repository at package installation time, febootstrap is also required. (c) Individual users may decide to switch to using the fixed appliance for performance reasons or whatever, but that doesn't mean that febootstrap can be made an optional package for all Debian users. I understand that 'Recommends' does not necessarily mean that the package won't be installed, but I still think there is scope here for an ordinary Debian user to break their libguestfs installation by uninstalling febootstrap, and in any case there is no real reason to uninstall febootstrap (see the next point). (d) I don't understand why you wouldn't want febootstrap on your institute[']s systems. It's a perfectly decent, fully free and interesting tool which takes a unique and innovative approach to the problem of packaging enormous virtual machines. It works on Debian and a variety of other distros. http://rwmj.wordpress.com/2009/10/22/supermin-appliance-now-in-febootstrap/ Rich. -- Richard Jones, Virtualization Group, Red Hat http://people.redhat.com/~rjones virt-df lists disk usage of guests without needing to install any software inside the virtual machine. Supports Linux and Windows. http://people.redhat.com/~rjones/virt-df/ -- To UNSUBSCRIBE, email to debian-bugs-dist-requ...@lists.debian.org with a subject of unsubscribe. Trouble? Contact listmas...@lists.debian.org
Bug#698771: Python classes should inherit from object
I believe this is fixed upstream in libguestfs: https://github.com/libguestfs/libguestfs/commit/abb1d466364f74adfcf3c985b92673285fd32e44 Also Dan Berrange made the equivalent change to libvirt, and I will also change hivex. Rich. -- Richard Jones, Virtualization Group, Red Hat http://people.redhat.com/~rjones virt-top is 'top' for virtual machines. Tiny program with many powerful monitoring features, net stats, disk stats, logging, etc. http://people.redhat.com/~rjones/virt-top -- To UNSUBSCRIBE, email to debian-bugs-dist-requ...@lists.debian.org with a subject of unsubscribe. Trouble? Contact listmas...@lists.debian.org
Bug#698771: Python classes should inherit from object
On Thu, Jan 24, 2013 at 03:57:26PM +0100, Hilko Bengen wrote: * Richard W.M. Jones: I believe this is fixed upstream in libguestfs: https://github.com/libguestfs/libguestfs/commit/abb1d466364f74adfcf3c985b92673285fd32e44 Thanks! It looks surprisingly simple. ;-) This should not break any APIs, should it? All the tests run :-) Looking at the documentation, I don't see how anything could get broken by this change unless code is doing really odd stuff in the guts of Python. Note also that this is the default in Python 3 -- in other words for Python 3 we haven't changed anything. Rich. -- Richard Jones, Virtualization Group, Red Hat http://people.redhat.com/~rjones virt-top is 'top' for virtual machines. Tiny program with many powerful monitoring features, net stats, disk stats, logging, etc. http://people.redhat.com/~rjones/virt-top -- To UNSUBSCRIBE, email to debian-bugs-dist-requ...@lists.debian.org with a subject of unsubscribe. Trouble? Contact listmas...@lists.debian.org
Bug#698771: Python classes should inherit from object
Do you have a suggested patch or can you show us how to do this? Rich. -- To UNSUBSCRIBE, email to debian-bugs-dist-requ...@lists.debian.org with a subject of unsubscribe. Trouble? Contact listmas...@lists.debian.org
Bug#646383: hivex: implicit pointer conversion
I have added a patch upstream which allows the extension to be compiled with Python 3: http://git.annexia.org/?p=hivex.git;a=commitdiff;h=207402a20ac8be74e767876f8ac93aba6292b2a9 Rich. -- Richard Jones, Virtualization Group, Red Hat http://people.redhat.com/~rjones libguestfs lets you edit virtual machines. Supports shell scripting, bindings from many languages. http://libguestfs.org -- To UNSUBSCRIBE, email to debian-bugs-dist-requ...@lists.debian.org with a subject of unsubscribe. Trouble? Contact listmas...@lists.debian.org
Bug#530427: [Pkg-libvirt-maintainers] About libguestfs in Debian
On Tue, Jul 19, 2011 at 06:34:15PM +0200, Hilko Bengen wrote: * Guido Günther: On Sun, Jul 17, 2011 at 02:18:11AM +0800, Liang Guo wrote: Hi, Guido, I found libguestfs git repository[1] has no commit since March, do you still working on it? What's the current state ? I think libguestfs is interesting, may I help to packaging it ? Yes, having it in Debian would be great. Hilko worked on febootstrap and on libguestfs recently: Here's a bit of an update: Except for some gnulib-releated test cases that don't work properly, I have been able to build packages of 1.10.x. Upstream generates some files in debian/ through autoconf/automake, this has been a bit of a mess to work around. We will drop the debian/ directory as soon as it's added to Debian. It's a pain for us to maintain this too. I will try to find some time tonight to straighten out what I have got so far and push that to the repository. Any help is welcome, of course. CC to Nikita Menkovich who has been working on the debian/ subdirectory upstream. Rich. -- Richard Jones, Virtualization Group, Red Hat http://people.redhat.com/~rjones virt-df lists disk usage of guests without needing to install any software inside the virtual machine. Supports Linux and Windows. http://et.redhat.com/~rjones/virt-df/ -- To UNSUBSCRIBE, email to debian-bugs-dist-requ...@lists.debian.org with a subject of unsubscribe. Trouble? Contact listmas...@lists.debian.org
Bug#530427: About libguestfs in Debian
On Sun, Jul 17, 2011 at 02:18:11AM +0800, Liang Guo wrote: Hi, Guido, I found libguestfs git repository[1] has no commit since March, do you still working on it? What's the current state ? I think libguestfs is interesting, may I help to packaging it ? Thank you! [1]http://anonscm.debian.org/gitweb/?p=pkg-libvirt/libguestfs.git;a=summary Hilko Bengen and Nikita Menkovich have both worked on Debian packaging, Nikita most recently. Also when 1.12 is released (most likely next week), I will be providing corresponding Debian and Ubuntu packages. Please feel welcome to join us on the libguestfs mailing list to discuss this further. Rich. -- Richard Jones, Virtualization Group, Red Hat http://people.redhat.com/~rjones virt-df lists disk usage of guests without needing to install any software inside the virtual machine. Supports Linux and Windows. http://et.redhat.com/~rjones/virt-df/ -- To UNSUBSCRIBE, email to debian-bugs-dist-requ...@lists.debian.org with a subject of unsubscribe. Trouble? Contact listmas...@lists.debian.org
Bug#627353: Please include insmod.static
Follow up to this, since it was posted to another mailing list: http://lists.debian.org/debian-devel/2011/05/msg00768.html Can anyone tell me where this klibc linked insmod is? busybox-static woule be a reasonable choice (1.5 MB), however the version compiled for Debian appears to lack insmod support. Rich. -- Richard Jones, Virtualization Group, Red Hat http://people.redhat.com/~rjones New in Fedora 11: Fedora Windows cross-compiler. Compile Windows programs, test, and build Windows installers. Over 70 libraries supprt'd http://fedoraproject.org/wiki/MinGW http://www.annexia.org/fedora_mingw -- To UNSUBSCRIBE, email to debian-bugs-dist-requ...@lists.debian.org with a subject of unsubscribe. Trouble? Contact listmas...@lists.debian.org
Bug#627353: Please include insmod.static
[Previous response appears to have been consumed in the aether] Can you clarify why it does? I do not really like wasting space on every Debian system without a good reason. Well I can tell you what we use it for in febootstrap: We boot the appliance using a custom initramfs which has a statically linked init (no libc6) and which has to load kernel modules in order to mount the real root filesystem. In order to load those, we need a statically linked insmod.static. However I understand that having insmod.static on every Debian machine just for this purpose is a waste of space. Can it be easily included in a sub-package? Rich. -- Richard Jones, Virtualization Group, Red Hat http://people.redhat.com/~rjones virt-top is 'top' for virtual machines. Tiny program with many powerful monitoring features, net stats, disk stats, logging, etc. http://et.redhat.com/~rjones/virt-top -- To UNSUBSCRIBE, email to debian-bugs-dist-requ...@lists.debian.org with a subject of unsubscribe. Trouble? Contact listmas...@lists.debian.org
Bug#627353: Please include insmod.static
Can you clarify why it does? I do not really like wasting space on every Debian system without a good reason. Well I can tell you what we use it for in febootstrap: We boot the appliance using a custom initramfs which has a statically linked init (no libc6) and which has to load kernel modules in order to mount the real root filesystem. In order to load those, we need a statically linked insmod.static. However I understand that having insmod.static on every Debian machine just for this purpose is a waste of space. Can it be easily included in a sub-package? Rich. -- Richard Jones, Virtualization Group, Red Hat http://people.redhat.com/~rjones New in Fedora 11: Fedora Windows cross-compiler. Compile Windows programs, test, and build Windows installers. Over 70 libraries supprt'd http://fedoraproject.org/wiki/MinGW http://www.annexia.org/fedora_mingw -- To UNSUBSCRIBE, email to debian-bugs-dist-requ...@lists.debian.org with a subject of unsubscribe. Trouble? Contact listmas...@lists.debian.org
Bug#606622: udev: please add a way to override warn_if_interactive
On Mon, Apr 04, 2011 at 01:11:59AM +0200, Marco d'Itri wrote: Richard, what should I do about this bug? I am happy with adding support in udev to detect libguestfs in some reasonable and unambiguous way to fix this, but I do not know how to do it. We're now setting RUNLEVEL in order to work around this bug. I still think the original code is wrong. Rich. -- Richard Jones, Virtualization Group, Red Hat http://people.redhat.com/~rjones virt-top is 'top' for virtual machines. Tiny program with many powerful monitoring features, net stats, disk stats, logging, etc. http://et.redhat.com/~rjones/virt-top -- To UNSUBSCRIBE, email to debian-bugs-dist-requ...@lists.debian.org with a subject of unsubscribe. Trouble? Contact listmas...@lists.debian.org
Bug#530427: ITP: libguestfs -- library for accessing and modifying guest disk images
Just a note that I'm providing Debian and Ubuntu packages here: http://libguestfs.org/download/binaries/debian-packages/ http://libguestfs.org/download/binaries/ubuntu-packages/ Rich. -- Richard Jones, Virtualization Group, Red Hat http://people.redhat.com/~rjones virt-df lists disk usage of guests without needing to install any software inside the virtual machine. Supports Linux and Windows. http://et.redhat.com/~rjones/virt-df/ -- To UNSUBSCRIBE, email to debian-bugs-dist-requ...@lists.debian.org with a subject of unsubscribe. Trouble? Contact listmas...@lists.debian.org
Bug#608103: guestfish: Broken dependencies on virt-edit
1.0.84 is very very old, and not supported upstream or advisable to use. We are providing libguestfs packages here which I think you should try instead: http://libguestfs.org/download/binaries/ Rich. -- Richard Jones, Virtualization Group, Red Hat http://people.redhat.com/~rjones New in Fedora 11: Fedora Windows cross-compiler. Compile Windows programs, test, and build Windows installers. Over 70 libraries supprt'd http://fedoraproject.org/wiki/MinGW http://www.annexia.org/fedora_mingw -- To UNSUBSCRIBE, email to debian-bugs-dist-requ...@lists.debian.org with a subject of unsubscribe. Trouble? Contact listmas...@lists.debian.org
Bug#606622: udev: please add a way to override warn_if_interactive
On Fri, Dec 10, 2010 at 05:25:40PM +0100, Marco d'Itri wrote: On Dec 10, Richard Jones rjo...@redhat.com wrote: When I boot a Debian appliance for libguestfs (http://libguestfs.org) the boot sits and waits for 60 seconds while it prints the large warn_if_interactive warning (when starting udev service). Please, I would rather detect what is different in the libguestfs environment than resort to ad-hoc hacks. Why are $RUNLEVEL and/or $PREVLEVEL not set? How does libguestfs boot guests? We're not running under init(8)/upstart/whatever. It's booting an initramfs with its own /init file, and udev start is called from that file. There is no runlevel and there are no init-/upstart-related environment variables set at all. Rich. -- Richard Jones, Virtualization Group, Red Hat http://people.redhat.com/~rjones virt-df lists disk usage of guests without needing to install any software inside the virtual machine. Supports Linux and Windows. http://et.redhat.com/~rjones/virt-df/ -- To UNSUBSCRIBE, email to debian-bugs-dist-requ...@lists.debian.org with a subject of unsubscribe. Trouble? Contact listmas...@lists.debian.org
Bug#606622: udev: please add a way to override warn_if_interactive
On Fri, Dec 10, 2010 at 05:48:52PM +0100, Marco d'Itri wrote: On Dec 10, Richard W.M. Jones rjo...@redhat.com wrote: There is no runlevel and there are no init-/upstart-related environment variables set at all. Maybe then this custom init should be modified to appear to be more like other init programs? Well we can export RUNLEVEL I guess, but I still think it is the warning which is wrong. No one else's udev init script does this, and Ubuntu appear to remove the warning altogether. Otherwise, please clarify how programs can detect being started in a libguestfs environment. There aren't any other programs that are started in this environment, except a handful that we start explicitly. Rich. -- Richard Jones, Virtualization Group, Red Hat http://people.redhat.com/~rjones virt-top is 'top' for virtual machines. Tiny program with many powerful monitoring features, net stats, disk stats, logging, etc. http://et.redhat.com/~rjones/virt-top -- To UNSUBSCRIBE, email to debian-bugs-dist-requ...@lists.debian.org with a subject of unsubscribe. Trouble? Contact listmas...@lists.debian.org
Bug#604735: new version of seabios resolves guest moved used index from ... when using virtio-blk
Package: seabios Version: 0.6.0+git20100710-2 Severity: important seabios 0.6.0+git20100710-2 is incompatible (sort of) with qemu 0.13 and virtio-blk. In some circumstances you will hit errors like: guest moved used index from 0 to [some number] I was hitting this, and was able to resolve this by building myself a deb of today's git (20101123). -- System Information: Debian Release: squeeze/sid APT prefers unstable APT policy: (500, 'unstable'), (500, 'stable'), (1, 'experimental') Architecture: amd64 (x86_64) Kernel: Linux 2.6.36-trunk-amd64 (SMP w/1 CPU core) Locale: LANG=en_GB.UTF-8, LC_CTYPE=en_GB.UTF-8 (charmap=UTF-8) Shell: /bin/sh linked to /bin/bash -- no debconf information -- Richard Jones, Virtualization Group, Red Hat http://people.redhat.com/~rjones New in Fedora 11: Fedora Windows cross-compiler. Compile Windows programs, test, and build Windows installers. Over 70 libraries supprt'd http://fedoraproject.org/wiki/MinGW http://www.annexia.org/fedora_mingw -- To UNSUBSCRIBE, email to debian-bugs-dist-requ...@lists.debian.org with a subject of unsubscribe. Trouble? Contact listmas...@lists.debian.org
Bug#530427: [Pkg-libvirt-maintainers] Upload to debian?
On Mon, Aug 23, 2010 at 01:52:13PM +0200, Guido Günther wrote: We basically need to split out the build of the appliance and make that downloadable from an external URI. Not much work but nobody got around to it so far. Another way would be to fix the supermin appliance build for Debian but that would involve a lot of dependencies to be installed on the VM host. In any case you'll want to wait until the latest raft of changes are upstream. See: https://www.redhat.com/archives/libguestfs/2010-August/msg00028.html Rich. -- Richard Jones, Virtualization Group, Red Hat http://people.redhat.com/~rjones virt-df lists disk usage of guests without needing to install any software inside the virtual machine. Supports Linux and Windows. http://et.redhat.com/~rjones/virt-df/ -- To UNSUBSCRIBE, email to debian-bugs-dist-requ...@lists.debian.org with a subject of unsubscribe. Trouble? Contact listmas...@lists.debian.org
Bug#530427: libguestfs Debian, an idea
On Thu, Jun 24, 2010 at 11:45:32AM +0800, Paul Wise wrote: How about the libguestfs package depends on each of the packages used inside the vm/appliance, then builds the vm on first install and updates the vm via file triggers whenever the packages outside the vm change? This means it doesn't need to access the network at build time or at install time and also gives free updates of the vm. This is kinda similar to how the initramfs is built on Debian. Could do this, but building the appliance from debs is quite slow (3-10 minutes), and really once you've built it you should test it (1-2 hours). When we build the Fedora package, we do all this work once on the builders, and can then be fairly certain that what we're about to ship to end users will actually work. What's the problem with running debirf on the Debian builders anyway? It shouldn't need any special permissions or network access. Rich. -- Richard Jones, Virtualization Group, Red Hat http://people.redhat.com/~rjones libguestfs lets you edit virtual machines. Supports shell scripting, bindings from many languages. http://et.redhat.com/~rjones/libguestfs/ See what it can do: http://et.redhat.com/~rjones/libguestfs/recipes.html -- To UNSUBSCRIBE, email to debian-bugs-dist-requ...@lists.debian.org with a subject of unsubscribe. Trouble? Contact listmas...@lists.debian.org
Bug#576250: febootstrap: segfaults and fails to complete bootstrapping
On Thu, Jun 17, 2010 at 03:04:26AM +0300, Fathi Boudra wrote: Anyway, a new upstream release will be appreciated. If you don't mind, I could upload a NMU if needed. Yes please. The version in Debian is ancient. Rich. -- Richard Jones, Virtualization Group, Red Hat http://people.redhat.com/~rjones New in Fedora 11: Fedora Windows cross-compiler. Compile Windows programs, test, and build Windows installers. Over 70 libraries supprt'd http://fedoraproject.org/wiki/MinGW http://www.annexia.org/fedora_mingw -- To UNSUBSCRIBE, email to debian-bugs-dist-requ...@lists.debian.org with a subject of unsubscribe. Trouble? Contact listmas...@lists.debian.org
Bug#576250: febootstrap: segfaults and fails to complete bootstrapping
On Thu, Jun 17, 2010 at 11:45:02AM +0300, Fathi Boudra wrote: Hi, I have uploaded febootstrap 2.7-1 to Debian. Unfortunately, this version still segfault: Complete! /usr/bin/fakeroot: line 1: 17946 Segmentation fault exit $RESULT These segfaults are almost certainly caused by glibc in the chroot not matching glibc outside the chroot. In practice when we do this in Fedora we ensure that the glibc versions are identical. This in turn may be caused because we aren't setting the correct LD_LIBRARY_PATH(s) up inside febootstrap, so that we're in effect mixing internal parts of the two different glibc versions. I haven't really had time to look into this in any detail. Rich. -- Richard Jones, Virtualization Group, Red Hat http://people.redhat.com/~rjones Read my programming blog: http://rwmj.wordpress.com Fedora now supports 80 OCaml packages (the OPEN alternative to F#) http://cocan.org/getting_started_with_ocaml_on_red_hat_and_fedora -- To UNSUBSCRIBE, email to debian-bugs-dist-requ...@lists.debian.org with a subject of unsubscribe. Trouble? Contact listmas...@lists.debian.org
Bug#578800: perl4caml: FTBFS with Perl 5.12.0: SVt_RV is gone
There is an upstream fix for this: http://git.annexia.org/?p=perl4caml.git;a=commitdiff;h=4cb12aa05bd5aa69ccfa1c6d41ab10bc79a3c3a3 Rich. -- Richard Jones, Virtualization Group, Red Hat http://people.redhat.com/~rjones virt-df lists disk usage of guests without needing to install any software inside the virtual machine. Supports Linux and Windows. http://et.redhat.com/~rjones/virt-df/ -- To UNSUBSCRIBE, email to debian-bugs-dist-requ...@lists.debian.org with a subject of unsubscribe. Trouble? Contact listmas...@lists.debian.org
Bug#530427: ITP: libguestfs -- library for accessing and modifying guest disk images
On Fri, Apr 09, 2010 at 04:39:55PM +0200, Olivier Berger wrote: Hi. On Sun, May 24, 2009 at 08:56:59PM +0100, Richard Jones wrote: * Package name: libguestfs Version : 1.0.31 Upstream Author : Richard W.M. Jones rjo...@redhat.com * URL : http://et.redhat.com/~rjones/libguestfs/ * License : LGPL Programming Lang: C, Shell script, various others Description : library for accessing and modifying guest disk images Any news about this packaging for Debian ? It's not obvious why it wouldn't have entered Debian for quite some time packages have been worked on :-/ There are i386 packages here: http://pkg-libvirt.alioth.debian.org/packages/ The last I heard Guido didn't want to bundle the appliance part with the Debian package. Rich. -- Richard Jones, Virtualization Group, Red Hat http://people.redhat.com/~rjones virt-top is 'top' for virtual machines. Tiny program with many powerful monitoring features, net stats, disk stats, logging, etc. http://et.redhat.com/~rjones/virt-top -- To UNSUBSCRIBE, email to debian-bugs-dist-requ...@lists.debian.org with a subject of unsubscribe. Trouble? Contact listmas...@lists.debian.org
Bug#576250: I get the same with both fedora-10 and fedora-12.
On Thu, Apr 08, 2010 at 01:39:00PM +1000, Erik de Castro Lopo wrote: Hi, I ran the commands (also tried with fedora-12 with same result): mkdir fed10-64 febootstrap fedora-10 fed10-64 http://mirror/pub/fedora/linux/releases/10/Everything/x86_64/os Which package is this? Segfaults are in any case caused by incompatibilities between glibc in the host and appliance. You have to only install an operating system which is compatible with the host OS. In Fedora we recommend only installing (eg) Fedora 12 on Fedora 12, because that way you know that glibc won't have problems. Rich. -- Richard Jones, Virtualization Group, Red Hat http://people.redhat.com/~rjones New in Fedora 11: Fedora Windows cross-compiler. Compile Windows programs, test, and build Windows installers. Over 70 libraries supprt'd http://fedoraproject.org/wiki/MinGW http://www.annexia.org/fedora_mingw -- To UNSUBSCRIBE, email to debian-bugs-dist-requ...@lists.debian.org with a subject of unsubscribe. Trouble? Contact listmas...@lists.debian.org
Bug#576250: I get the same with both fedora-10 and fedora-12.
On Thu, Apr 08, 2010 at 08:17:16PM +1000, Erik de Castro Lopo wrote: Which package is this? ii febootstrap 2.1-4tool for bootstrapping a Fedora system (like Debian debootstrap) This is kind of old, I really need to update this to the latest version. So why is febootstrap even made available to be installed under debian? It does in fact work, but not in all cases. Rich. -- Richard Jones, Virtualization Group, Red Hat http://people.redhat.com/~rjones libguestfs lets you edit virtual machines. Supports shell scripting, bindings from many languages. http://et.redhat.com/~rjones/libguestfs/ See what it can do: http://et.redhat.com/~rjones/libguestfs/recipes.html -- To UNSUBSCRIBE, email to debian-bugs-dist-requ...@lists.debian.org with a subject of unsubscribe. Trouble? Contact listmas...@lists.debian.org
Bug#552394: febootstrap fails during rpm_check_debug
On Tue, Oct 27, 2009 at 05:34:27PM +0530, Ritesh Raj Sarraf wrote: On Tuesday 27 Oct 2009 15:44:01 you wrote: See this blog entry: http://www.lucas-nussbaum.net/blog/?p=385 You need an even newer version of rpm from Debian/experimental. The package needs. It should then depend on the version in Debian/experimental. That's debatable: The repository that you are trying to install needs that updated RPM. febootstrap itself doesn't need it. If there's some Debian policy around this please let me know. Rich. -- Richard Jones, Virtualization Group, Red Hat http://people.redhat.com/~rjones libguestfs lets you edit virtual machines. Supports shell scripting, bindings from many languages. http://et.redhat.com/~rjones/libguestfs/ See what it can do: http://et.redhat.com/~rjones/libguestfs/recipes.html -- To UNSUBSCRIBE, email to debian-bugs-dist-requ...@lists.debian.org with a subject of unsubscribe. Trouble? Contact listmas...@lists.debian.org
Bug#552394: febootstrap fails during rpm_check_debug
On Tue, Oct 27, 2009 at 09:58:35AM +0530, Ritesh Raj Sarraf wrote: On Monday 26 Oct 2009 22:26:38 Richard W.M. Jones wrote: I think this is because of an old version of RPM. Which version of RPM do you have? r...@champaran:~ $ rpm --version RPM version 4.7.0 See this blog entry: http://www.lucas-nussbaum.net/blog/?p=385 You need an even newer version of rpm from Debian/experimental. Rich. -- Richard Jones, Virtualization Group, Red Hat http://people.redhat.com/~rjones New in Fedora 11: Fedora Windows cross-compiler. Compile Windows programs, test, and build Windows installers. Over 70 libraries supprt'd http://fedoraproject.org/wiki/MinGW http://www.annexia.org/fedora_mingw -- To UNSUBSCRIBE, email to debian-bugs-dist-requ...@lists.debian.org with a subject of unsubscribe. Trouble? Contact listmas...@lists.debian.org
Bug#552394: febootstrap fails during rpm_check_debug
I think this is because of an old version of RPM. Which version of RPM do you have? Rich. -- Richard Jones, Virtualization Group, Red Hat http://people.redhat.com/~rjones virt-df lists disk usage of guests without needing to install any software inside the virtual machine. Supports Linux and Windows. http://et.redhat.com/~rjones/virt-df/ -- To UNSUBSCRIBE, email to debian-bugs-dist-requ...@lists.debian.org with a subject of unsubscribe. Trouble? Contact listmas...@lists.debian.org
Bug#550118: RFP: vhostmd -- virtual host metrics daemon
Package: wnpp Severity: wishlist http://gitorious.org/vhostmd/pages/Home http://gitorious.org/vhostmd vhostmd (the Virtual Host Metrics Daemon) allows virtual machines to see limited information about the host they are running on, in a secure and highly configurable way. -- Richard Jones, Emerging Technologies, Red Hat http://et.redhat.com/~rjones Read my programming blog: http://rwmj.wordpress.com -- To UNSUBSCRIBE, email to debian-bugs-dist-requ...@lists.debian.org with a subject of unsubscribe. Trouble? Contact listmas...@lists.debian.org
Bug#536214: febootstrap: UnicodeDecodeError when rpm_check_debug is run
I opened an upstream bug here: https://bugzilla.redhat.com/show_bug.cgi?id=510194 Is it possible you can repeat this and show the full messages? I'm interested to see what packages it is trying to install. I repeated your test using your LANG and LC_CTYPE settings, but with a Fedora host, and it worked OK for me. Rich. -- Richard Jones, Emerging Technologies, Red Hat http://et.redhat.com/~rjones Read my programming blog: http://rwmj.wordpress.com Fedora now supports 75 OCaml packages (the OPEN alternative to F#) http://cocan.org/getting_started_with_ocaml_on_red_hat_and_fedora -- To UNSUBSCRIBE, email to debian-bugs-dist-requ...@lists.debian.org with a subject of unsubscribe. Trouble? Contact listmas...@lists.debian.org
Bug#536214: febootstrap: UnicodeDecodeError when rpm_check_debug is run
On Wed, Jul 08, 2009 at 01:57:02PM +0200, Arthur Lutz wrote: ERREUR de résolution de dépendance par rpm_check_debug : Traceback (most recent call last): File /usr/bin/yum, line 29, in module yummain.user_main(sys.argv[1:], exit_code=True) File /usr/share/yum-cli/yummain.py, line 309, in user_main errcode = main(args) File /usr/share/yum-cli/yummain.py, line 261, in main return_code = base.doTransaction() File /usr/share/yum-cli/cli.py, line 424, in doTransaction print to_utf8(msg) File /usr/lib/python2.5/codecs.py, line 303, in write data, consumed = self.encode(object, self.errors) UnicodeDecodeError: 'ascii' codec can't decode byte 0xc3 in position 25: ordinal not in range(128) Yes, this really does look like a yum bug to me ... Rich. -- Richard Jones, Emerging Technologies, Red Hat http://et.redhat.com/~rjones libguestfs lets you edit virtual machines. Supports shell scripting, bindings from many languages. http://et.redhat.com/~rjones/libguestfs/ See what it can do: http://et.redhat.com/~rjones/libguestfs/recipes.html -- To UNSUBSCRIBE, email to debian-bugs-dist-requ...@lists.debian.org with a subject of unsubscribe. Trouble? Contact listmas...@lists.debian.org
Bug#365349: Also virt-mem
Also virt-mem needs kernel debuginfo-type information. Is there any progress on providing this for Debian? Rich. -- Richard Jones, Emerging Technologies, Red Hat http://et.redhat.com/~rjones Read my OCaml programming blog: http://camltastic.blogspot.com/ Fedora now supports 60 OCaml packages (the OPEN alternative to F#) http://cocan.org/getting_started_with_ocaml_on_red_hat_and_fedora -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]
Bug#453955: source for prelimenary virt-top debian packages
On Sat, Jun 28, 2008 at 12:23:25PM +0200, Guido Günther wrote: Hi, the source for prelimenary Debian packages is available here: http://git.debian.org/?p=users/agx/virt-top.git Those will be uploaded once we've moved them over to pkg-ocaml. -- Guido -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]
Bug#402249: [Pkg-xen-devel] Bug#402249: [PATCH] include necessary headers to build libvirt with xen support
Bastian Blank wrote: On Sat, Dec 15, 2007 at 07:40:07PM +0100, Guido Guenther wrote: for your convenience I attach Richard's patch directly. It simply installs the necessary headers so we're able to enable xen support in libvirt for the default builds. Please apply. libvirt redefines several parts of this headers. So they can include the missing pieces also. We include redefinitions in a very few places, particularly where the Xen headers used to include definitions which have since been removed. Last week a bug was raised about this, I'm not sure if this is what you're referring to: https://www.redhat.com/archives/libvir-list/2007-December/msg00251.html The greater point is that the headers which this patch includes are part of the public Xen API, and these header files are normally exported. Except that in both Debian and Fedora, because of the specific way that Debian and Fedora packaging call into a part of the Xen system of makefiles, they don't get exported. Specifically: $(MAKE) -C $(SOURCE_DIR)/tools install [...] Debian only do the 'make install' in a subdirectory. If Debian did a normal 'make install', these headers would be installed, but there are other drawbacks to doing that, so exporting the headers by hand is the next best thing. Rich. -- Emerging Technologies, Red Hat - http://et.redhat.com/~rjones/ Registered Address: Red Hat UK Ltd, Amberley Place, 107-111 Peascod Street, Windsor, Berkshire, SL4 1TE, United Kingdom. Registered in England and Wales under Company Registration No. 03798903 smime.p7s Description: S/MIME Cryptographic Signature
Bug#453953: Need to ocaml-csv
You don't necessarily need ocaml-csv to compile virt-top. If you don't have it you'll lose a bit of functionality (basically all the --csv command line options) but it will still compile and run. Another thing to note is the Debian OCaml packaging policy: http://pkg-ocaml-maint.alioth.debian.org/ocaml_packaging_policy.html/index.html smime.p7s Description: S/MIME Cryptographic Signature
Bug#384300: bug #384300 - libvirt package for Debian
This is a libvirt package for Debian. It's derived from the Ubuntu libvirt package, rebuilt for Debian testing. I'm not the greatest Debian packager in the world, so any comments / feedback are welcome. The previous patch to the base Xen package is also required before this can be built. Source: http://et.redhat.com/~rjones/debian-libvirt/libvirt_0.3.3.orig.tar.gz http://et.redhat.com/~rjones/debian-libvirt/libvirt_0.3.3-1_i386.changes http://et.redhat.com/~rjones/debian-libvirt/libvirt_0.3.3-1.dsc http://et.redhat.com/~rjones/debian-libvirt/libvirt_0.3.3-1.diff.gz Binary for i386: http://et.redhat.com/~rjones/debian-libvirt/libvirt-bin_0.3.3-1_i386.deb http://et.redhat.com/~rjones/debian-libvirt/libvirt-dev_0.3.3-1_i386.deb http://et.redhat.com/~rjones/debian-libvirt/libvirt0_0.3.3-1_i386.deb http://et.redhat.com/~rjones/debian-libvirt/python-libvirt_0.3.3-1_i386.deb There is a Debian ITP for libvirt here: http://bugs.debian.org/cgi-bin/bugreport.cgi?bug=384300 Rich. -- Emerging Technologies, Red Hat - http://et.redhat.com/~rjones/ Registered Address: Red Hat UK Ltd, Amberley Place, 107-111 Peascod Street, Windsor, Berkshire, SL4 1TE, United Kingdom. Registered in England and Wales under Company Registration No. 03798903 smime.p7s Description: S/MIME Cryptographic Signature
Bug#384443: bug #384443 - packages for virt-install and virt-manager
This is a Debian package for virt-install (a dependency of virt-manager). It is both a standalone command line tool for installing new virtual guests, and a library of Python functions used by virt-manager for the same task. Source: http://et.redhat.com/~rjones/debian-libvirt/virt-install_0.300.1-9.diff.gz http://et.redhat.com/~rjones/debian-libvirt/virt-install_0.300.1-9.dsc http://et.redhat.com/~rjones/debian-libvirt/virt-install_0.300.1.orig.tar.gz i386 binary: http://et.redhat.com/~rjones/debian-libvirt/virt-install_0.300.1-9_i386.changes http://et.redhat.com/~rjones/debian-libvirt/virt-install_0.300.1-9_i386.deb Next up is gtk-vnc (a dependency of virt-manager). This is a VNC client widget. Source: http://et.redhat.com/~rjones/debian-libvirt/gtk-vnc_0.2.0-1.diff.gz http://et.redhat.com/~rjones/debian-libvirt/gtk-vnc_0.2.0-1.dsc http://et.redhat.com/~rjones/debian-libvirt/gtk-vnc_0.2.0.orig.tar.gz i386 binaries: http://et.redhat.com/~rjones/debian-libvirt/gtk-vnc_0.2.0-1_i386.changes http://et.redhat.com/~rjones/debian-libvirt/libgtk-vnc-dev_0.2.0-1_i386.deb http://et.redhat.com/~rjones/debian-libvirt/libgtk-vnc0_0.2.0-1_i386.deb http://et.redhat.com/~rjones/debian-libvirt/python-gtk-vnc_0.2.0-1_i386.deb Finally, a Debian package for virt-manager, a graphical tool for managing virtual guests. Source: http://et.redhat.com/~rjones/debian-libvirt/virt-manager_0.5.2-7.diff.gz http://et.redhat.com/~rjones/debian-libvirt/virt-manager_0.5.2-7.dsc http://et.redhat.com/~rjones/debian-libvirt/virt-manager_0.5.2.orig.tar.gz i386 binary: http://et.redhat.com/~rjones/debian-libvirt/virt-manager_0.5.2-7_i386.changes http://et.redhat.com/~rjones/debian-libvirt/virt-manager_0.5.2-7_i386.deb There was an ITP for virt-manager here: http://bugs.debian.org/cgi-bin/bugreport.cgi?bug=384443 Rich. -- Emerging Technologies, Red Hat - http://et.redhat.com/~rjones/ Registered Address: Red Hat UK Ltd, Amberley Place, 107-111 Peascod Street, Windsor, Berkshire, SL4 1TE, United Kingdom. Registered in England and Wales under Company Registration No. 03798903 smime.p7s Description: S/MIME Cryptographic Signature
Bug#384443: [Libvir] bug #384443 - packages for virt-install and virt-manager
Sorry, missed one. virt-viewer is a soft dependency for virt-install in that it's required for graphical installs. http://et.redhat.com/~rjones/debian-libvirt/virt-viewer_0.0.2-1.diff.gz http://et.redhat.com/~rjones/debian-libvirt/virt-viewer_0.0.2-1.dsc http://et.redhat.com/~rjones/debian-libvirt/virt-viewer_0.0.2-1_i386.changes http://et.redhat.com/~rjones/debian-libvirt/virt-viewer_0.0.2-1_i386.deb http://et.redhat.com/~rjones/debian-libvirt/virt-viewer_0.0.2.orig.tar.gz Rich. -- Emerging Technologies, Red Hat - http://et.redhat.com/~rjones/ Registered Address: Red Hat UK Ltd, Amberley Place, 107-111 Peascod Street, Windsor, Berkshire, SL4 1TE, United Kingdom. Registered in England and Wales under Company Registration No. 03798903 smime.p7s Description: S/MIME Cryptographic Signature
Bug#384300: Update?
Since xen-utils contains the required libraries, is it now possible to package libvirt? Rich. -- Emerging Technologies, Red Hat - http://et.redhat.com/~rjones/ Registered Address: Red Hat UK Ltd, Amberley Place, 107-111 Peascod Street, Windsor, Berkshire, SL4 1TE, United Kingdom. Registered in England and Wales under Company Registration No. 03798903 smime.p7s Description: S/MIME Cryptographic Signature
