inspector: Determine guest architecture.

This commit adds kernel and userspace architecture detection
(fully for Linux, partially for Windows).  It adds an architecture
for each kernel detected, and an architecture for each root (ie.
userspace) found.

3 files changed, 113 insertions, 21 deletions

diff --git a/TODO b/TODO
index 291d220a..dbfc3c88 100644
--- a/TODO
+++ b/TODO
@@ -182,22 +182,6 @@ Swap space
 
 Allow swap space from the guest to be used.  Is it a good idea?
 
-Query guest architecture
-------------------------
-
-Need a way to query a binary or library file for its architecture.
-Using objdump or readelf?
-What about non-ELF files (eg. Windows, BSD).
-
-To do this properly requires some serious logic, eg. to cover Linux
-and Windows we'd need objdump and i686-pc-mingw32-objdump, and more to
-cover a.out, COFF and 64 bit Windows.  Therefore this cannot be done
-inside the daemon, and should be done by a separate, external program
-similar to virt-inspector.
-
-Probably we should go all the way and have virt-inspector able to
-determine kernel and userspace architectures of guests.
-
 Other initrd-* commands
 -----------------------
 
diff --git a/inspector/virt-inspector.pl b/inspector/virt-inspector.pl
index 9148a486..d2acf062 100755
--- a/inspector/virt-inspector.pl
+++ b/inspector/virt-inspector.pl
@@ -357,6 +357,7 @@ sub output_text_os
 
     print $os->{os}, " " if exists $os->{os};
     print $os->{distro}, " " if exists $os->{distro};
+    print $os->{arch}, " " if exists $os->{arch};
     print $os->{major_version} if exists $os->{major_version};
     print ".", $os->{minor_version} if exists $os->{minor_version};
     print " ";
@@ -415,7 +416,7 @@ sub output_text_os
     print __"  Kernels:\n";
     my @kernels = @{$os->{kernels}};
     foreach (@kernels) {
-	print "    $_->{version}\n";
+	print "    $_->{version} ($_->{arch})\n";
 	my @modules = @{$_->{modules}};
 	foreach (@modules) {
 	    print "      $_\n";
@@ -450,6 +451,7 @@ sub output_xml_os
 
     foreach ( [ "name" => "os" ],
               [ "distro" => "distro" ],
+              [ "arch" => "arch" ],
               [ "major_version" => "major_version" ],
               [ "minor_version" => "minor_version" ],
               [ "package_format" => "package_format" ],
@@ -531,7 +533,9 @@ sub output_xml_os
     $xml->startTag("kernels");
     my @kernels = @{$os->{kernels}};
     foreach (@kernels) {
-        $xml->startTag("kernel", "version" => $_->{version});
+        $xml->startTag("kernel",
+		       "version" => $_->{version},
+		       "arch" => $_->{arch});
         $xml->startTag("modules");
 	my @modules = @{$_->{modules}};
 	foreach (@modules) {
@@ -591,6 +595,8 @@ sub output_query
     output_query_xen_domU_kernel ();
     output_query_xen_pv_drivers ();
     output_query_virtio_drivers ();
+    output_query_kernel_arch ();
+    output_query_userspace_arch ();
 }
 
 =item windows=(yes|no)
@@ -763,6 +769,50 @@ sub output_query_virtio_drivers
     print "virtio_drivers=no\n";
 }
 
+=item userspace_arch=(x86_64|...)
+
+Print the architecture of userspace.
+
+NB. For multi-boot VMs this can print several lines.
+
+=cut
+
+sub output_query_userspace_arch
+{
+    my %arches;
+
+    foreach my $os (keys %$oses) {
+	$arches{$oses->{$os}->{arch}} = 1 if exists $oses->{$os}->{arch};
+    }
+
+    foreach (sort keys %arches) {
+	print "userspace_arch=$_\n";
+    }
+}
+
+=item kernel_arch=(x86_64|...)
+
+Print the architecture of the kernel.
+
+NB. For multi-boot VMs this can print several lines.
+
+=cut
+
+sub output_query_kernel_arch
+{
+    my %arches;
+
+    foreach my $os (keys %$oses) {
+	foreach my $kernel (@{$oses->{$os}->{kernels}}) {
+	    $arches{$kernel->{arch}} = 1 if exists $kernel->{arch};
+	}
+    }
+
+    foreach (sort keys %arches) {
+	print "kernel_arch=$_\n";
+    }
+}
+
 =back
 
 =head1 SEE ALSO
diff --git a/perl/lib/Sys/Guestfs/Lib.pm b/perl/lib/Sys/Guestfs/Lib.pm
index 081b4b9b..1e5638e3 100644
--- a/perl/lib/Sys/Guestfs/Lib.pm
+++ b/perl/lib/Sys/Guestfs/Lib.pm
@@ -876,6 +876,17 @@ sub _check_linux_root
 	}
 	$r->{fstab} = \@fstab if @fstab;
     }
+
+    # Determine the architecture of this root.
+    my $arch;
+    foreach ("/bin/bash", "/bin/ls", "/bin/echo", "/bin/rm", "/bin/sh") {
+	if ($g->is_file ($_)) {
+	    $arch = file_architecture ($g, $_);
+	    last;
+	}
+    }
+
+    $r->{arch} = $arch if defined $arch;
 }
 
 # We only support NT.  The control file /boot.ini contains a list of
@@ -914,13 +925,30 @@ sub _check_windows_root
 
 	if (defined $systemroot) {
 	    $r->{systemroot} = resolve_windows_path ($g, "/$systemroot");
-	    if (defined $r->{systemroot} && $use_windows_registry) {
-		_check_windows_registry ($g, $r, $r->{systemroot});
+	    if (defined $r->{systemroot}) {
+		_check_windows_arch ($g, $r, $r->{systemroot});
+		if ($use_windows_registry) {
+		    _check_windows_registry ($g, $r, $r->{systemroot});
+		}
 	    }
 	}
     }
 }
 
+# Find Windows userspace arch.
+
+sub _check_windows_arch
+{
+    local $_;
+    my $g = shift;
+    my $r = shift;
+    my $systemroot = shift;
+
+    my $cmd_exe =
+	resolve_windows_path ($g, $r->{systemroot} . "/system32/cmd.exe");
+    $r->{arch} = file_architecture ($g, $cmd_exe) if $cmd_exe;
+}
+
 sub _check_windows_registry
 {
     local $_;
@@ -1034,6 +1062,10 @@ The C<\%os> hash contains the following keys (any can be omitted):
 
 Operating system type, eg. "linux", "windows".
 
+=item arch
+
+Operating system userspace architecture, eg. "i386", "x86_64".
+
 =item distro
 
 Operating system distribution, eg. "debian".
@@ -1118,6 +1150,7 @@ sub _get_os_version
         if exists $r->{root}->{package_format};
     $r->{package_management} = $r->{root}->{package_management}
         if exists $r->{root}->{package_management};
+    $r->{arch} = $r->{root}->{arch} if exists $r->{root}->{arch};
 }
 
 sub _assign_mount_points
@@ -1267,6 +1300,24 @@ List of applications.
 
 List of kernels.
 
+This is a hash of kernel version =E<gt> a hash with the following keys:
+
+=over 4
+
+=item version
+
+Kernel version.
+
+=item arch
+
+Kernel architecture (eg. C<x86-64>).
+
+=item modules
+
+List of modules.
+
+=back
+
 =item modprobe_aliases
 
 (For Linux VMs).
@@ -1359,14 +1410,21 @@ sub _check_for_kernels
 
 		# List modules.
 		my @modules;
-		foreach ($g->find ("/lib/modules/$_")) {
+		my $any_module;
+		my $prefix = "/lib/modules/$_";
+		foreach ($g->find ($prefix)) {
 		    if (m,/([^/]+)\.ko$, || m,([^/]+)\.o$,) {
+			$any_module = "$prefix$_" unless defined $any_module;
 			push @modules, $1;
 		    }
 		}
 
 		$kernel{modules} = \@modules;
 
+		# Determine kernel architecture by looking at the arch
+		# of any kernel module.
+		$kernel{arch} = file_architecture ($g, $any_module);
+
 		push @kernels, \%kernel;
 	    }
 	}