search for: _remap_block_devices

...SE changes will then be limited in scope to only what is required for SUSE support. This series of patches accomplishes the following: - Renames RedHat.pm to Linux.pm - Modifies Linux.pm to be more generic, and ready for additional distribution support (such as SUSE) - Adds functionality to _remap_block_devices, to ensure root and resume devices are updated in both grub legacy and grub2 environments - Allows for qxl or cirrus settings in LibVirtTarget.pm (based on passing guestcaps->{display} from the conversion. I'll submit the SUSE-specific changes after we work through any issues raised...

...ecommended > by Matt Booth in: > > https://www.redhat.com/archives/libguestfs/2013-September/msg00076.html > > Some aspects of this patch still need additional testing, and a couple of > changes are not foolproof (such as the lack of grub2 support in the menu.lst > changes in _remap_block_devices). However, it is complete enough that I'd > like some feedback before continuing. > > Note - This patch alone is not enough to support SUSE guests, as changes to > virt-v2v.db are also required. After these changes are flushed out, I'll post > all the patches in one thread....

...++++- 1 file changed, 17 insertions(+), 1 deletion(-) diff --git a/lib/Sys/VirtConvert/Converter/RedHat.pm b/lib/Sys/VirtConvert/Converter/RedHat.pm index 36345e7..fc4f5f0 100644 --- a/lib/Sys/VirtConvert/Converter/RedHat.pm +++ b/lib/Sys/VirtConvert/Converter/RedHat.pm @@ -2234,6 +2234,10 @@ sub _remap_block_devices # Fedora has used libata since FC7, which is long out of support. We assume # that all Fedora distributions in use use libata. + # Create a hash to track any hd->sd conversions, as any references to + # hd devices (in fstab, menu.lst, etc) will need to be converted later. +...

During the conversion process, hd* block devices under Xen are mapped to sd* devices for use under guestfs. These sd devices are later remapped to vd* (or sd/hd) devices, for future use under kvm. The current code fails to track the original hd* device names. This causes subsequent remappings of the block devices (in such places as /etc/fstab, and /boot/grub/menu.lst), to skip the hd* devices,

...------------ 1 files changed, 0 insertions(+), 28 deletions(-) diff --git a/lib/Sys/VirtV2V/Converter/Linux.pm b/lib/Sys/VirtV2V/Converter/Linux.pm index 87eeeb3..1186430 100644 --- a/lib/Sys/VirtV2V/Converter/Linux.pm +++ b/lib/Sys/VirtV2V/Converter/Linux.pm @@ -133,14 +133,6 @@ sub convert sub _remap_block_devices { my ($guestos, $dom, $desc, $virtio) = @_; - die("remap_block_devices called without guestos argument") - unless defined($guestos); - die("remap_block_devices called without dom argument") - unless defined($dom); - die("remap_block_devices call...

...when I found this patchset :/ Rather than redo it, I've kept what I'd already done and rebased on top of it. I also did some testing and fixed a couple of bugs: "/files$grub->{grub_conf}*/kernel/root" should have been "/files $grub->{grub_conf}/*/kernel/root" in _remap_block_devices. It turns out that the release suffix used by RHEL 4 and 5 really is 'xen', not '-xen'. Note that this is the release, not the package name, which is what confused me when I originally reviewed it. I've removed this change. Can you please pull the latest from git and rebase an...

...'Failed to install user-custom packages'); } @@ -625,8 +726,9 @@ sub convert my $remove_serial_console = exists($options->{NO_SERIAL_CONSOLE}); _configure_console($g, $grub, $remove_serial_console); - _configure_display_driver($g, $root, $config, $meta, $grub); - _remap_block_devices($meta, $virtio, $g, $root); + my $driver = _get_display_driver($g, $root); + _configure_display_driver($g, $root, $config, $meta, $grub, $driver); + _remap_block_devices($meta, $virtio, $g, $root, $grub); _configure_kernel_modules($g, $virtio); _configure_boot($kernel, $virtio, $...

...;No grub configuration found') unless defined($grub); @@ -625,8 +719,9 @@ sub convert my $remove_serial_console = exists($options->{NO_SERIAL_CONSOLE}); _configure_console($g, $grub, $remove_serial_console); - _configure_display_driver($g, $root, $config, $meta, $grub); - _remap_block_devices($meta, $virtio, $g, $root); + my $driver = _get_display_driver($g, $root); + _configure_display_driver($g, $root, $config, $meta, $grub, $driver); + _remap_block_devices($meta, $virtio, $g, $root, $grub); _configure_kernel_modules($g, $virtio); _configure_boot($kernel, $virtio, $...

...s($initrd, followsymlinks=>1)); + # Catch invalid paths by ensuring first character is '/' + if ($initrd =~ /^\//) { + return $initrd if $g->is_file_opts($initrd, followsymlinks=>1); + } } } @@ -2672,13 +2675,13 @@ sub _remap_block_devices # Add standard configuration files to the checklist push (@checklist, '/files/etc/fstab/*/spec'); + push (@checklist, '/files/boot/*/device.map/*'. + '[label() != "#comment"]'); # Add grub or grub2 file...

On Friday, October 04, 2013 09:38:58 AM Matthew Booth wrote: > It's specifically an error if we're attempting to configure virtio, and > there's no detected virtio kernel. It shouldn't have been possible to > get here in that state, hence it's a programmer error. The code below > attempts to install *any* kernel in the case that we aren't configuring >

If a volume is not an exact multiple of 512 bytes, qemu-img will report its size rounded down to a 512 byte boundary. However, when copying, the file is still read until EOF, which will return more data than was expected. This change prevents that causing a failure in itself. The situation is still not resolved, however, as there are still situations where this will cause a failure. For example,

...commands, and the dependencies cannot be satisfied. This is why I had to add the -base package to $kernel somehow... (Note - This doesn't effect zypper installs at all as it will satisfy dependencies automatically.) I'll think about it some more... > > @@ -2235,6 +2541,10 @@ sub _remap_block_devices ... > I've had a change to think about this change properly, and I'm pretty > sure it's not required. The reason is that the only place %idemap is > modified is in the section guarded: > > if ($libata) { > ... > } > > If the guest is using libata it will not...

...lls at all as it will satisfy dependencies automatically.) > > I'll think about it some more... You can still make $kernel a list, you just need to ensure it's always a list. The updates to the other installation methods should be trivial. > > > @@ -2235,6 +2541,10 @@ sub _remap_block_devices > ... > > I've had a change to think about this change properly, and I'm pretty > > sure it's not required. The reason is that the only place %idemap is > > modified is in the section guarded: > > > > if ($libata) { > > ... > > } > >...

...convert called without devices argument") unless defined($devices); # Un-configure HV specific attributes which don't require a direct # replacement @@ -117,7 +118,7 @@ sub convert # Configure the rest of the system _configure_display_driver($guestos, $virtio); - _remap_block_devices($guestos, $dom, $desc, $virtio); + $guestos->remap_block_devices($devices, $virtio); _configure_kernel_modules($guestos, $desc, $virtio); _configure_boot($guestos, $kernel, $virtio); @@ -130,32 +131,6 @@ sub convert return \%guestcaps; } -sub _remap_block_devices -{ - m...

...onversion. This is first approach recommended by Matt Booth in: https://www.redhat.com/archives/libguestfs/2013-September/msg00076.html Some aspects of this patch still need additional testing, and a couple of changes are not foolproof (such as the lack of grub2 support in the menu.lst changes in _remap_block_devices). However, it is complete enough that I'd like some feedback before continuing. Note - This patch alone is not enough to support SUSE guests, as changes to virt-v2v.db are also required. After these changes are flushed out, I'll post all the patches in one thread. -Mike --- lib/Sys/Virt...

The following series of patches adds support for converting SUSE guests through virt-v2v. These changes should not impact non-SUSE guest conversions. Mike Latimer (4): Add perl-Bootloader support to grub packages Add conversion support for SUSE guests Add SUSE to capabilities db and conf Add SUSE support documentation lib/Sys/VirtConvert/Converter/Linux.pm | 480

...rect # replacement - _unconfigure_hv($g, $desc); + _unconfigure_hv($g, $root, $desc); # Try to install the virtio capability my $virtio = _install_capability('virtio', $g, $config, $meta, $desc); @@ -125,7 +130,7 @@ sub convert _configure_display_driver($g); _remap_block_devices($meta, $virtio, $g, $desc); _configure_kernel_modules($g, $desc, $virtio, $modpath); - _configure_boot($kernel, $virtio, $g, $desc); + _configure_boot($kernel, $virtio, $g, $root, $desc); my %guestcaps; @@ -850,18 +855,18 @@ sub _configure_kernel sub _configure_boot { - m...

...); > + } > + > + # Configure the rest of the system > + my $remove_serial_console = exists($options->{NO_SERIAL_CONSOLE}); > + _configure_console($g, $grub, $remove_serial_console); > + > + _configure_display_driver($g, $root, $config, $meta, $grub); > + _remap_block_devices($meta, $virtio, $g, $root, $grub); You've added $grub here. See note about this later. > +# Configure a console on ttyS0. Make sure existing console references use it. > +# N.B. Note that the RHEL 6 xen guest kernel presents a console device called > +# /dev/hvc0, where previous xen...

This is a new conversion module to convert SUSE Linux and openSUSE guests. The converter is based on the RedHat module, and should offer the same functionality on both SUSE and RedHat hosts. There are a few additional messages in this module, such as reporting of packages when installing through zypper or the local virt-v2v repo. These messages don't necessarily flow unless verbose switches

...led to install user-custom packages'); + } + + # Configure the rest of the system + my $remove_serial_console = exists($options->{NO_SERIAL_CONSOLE}); + _configure_console($g, $grub, $remove_serial_console); + + _configure_display_driver($g, $root, $config, $meta, $grub); + _remap_block_devices($meta, $virtio, $g, $root, $grub); + _configure_kernel_modules($g, $virtio); + _configure_boot($kernel, $virtio, $g, $root, $grub); + + my %guestcaps; + + $guestcaps{block} = $virtio == 1 ? 'virtio' : 'ide'; + $guestcaps{net} = $virtio == 1 ? 'virtio' : '...