On 8/3/11 5:52 AM, Aneesh Kumar K.V wrote:
On Tue,  2 Aug 2011 21:46:29 +0300, Sasha Levin<[email protected]>  wrote:
This patch adds 2 new flags:

  --rootfs [path] - Specifies a path to use as rootfs. The path will be mounted
using virtio-9p and booted from.
Easiest way to test it is to mount the sample image we recommend with kvm tool
(http://wiki.qemu.org/download/linux-0.2.img.bz2) and mounting it somewhere.
*PLEASE DO NOT TRY BOOTING FROM YOUR REAL ROOT AKA '/', EVEN WITH READ ONLY
SET* - Lets make sure the code is bulletproof before we delete your rootfs :)

  --binsh - This flag will boot into simple /bin/sh shell instead of going
through /sbin/init. This is usefull for more complex distribution rootfs
where we have to deal with /etc/fstab and such.

Signed-off-by: Sasha Levin<[email protected]>
Reviewed-by: Aneesh Kumar K.V<[email protected]>

---
  tools/kvm/builtin-run.c |   18 +++++++++++++++++-
  1 files changed, 17 insertions(+), 1 deletions(-)

diff --git a/tools/kvm/builtin-run.c b/tools/kvm/builtin-run.c
index 2e04265..78680ec 100644
--- a/tools/kvm/builtin-run.c
+++ b/tools/kvm/builtin-run.c
@@ -76,11 +76,13 @@ static const char *guest_mac;
  static const char *host_mac;
  static const char *script;
  static const char *guest_name;
+static const char *rootfs;
  static bool single_step;
  static bool readonly_image[MAX_DISK_IMAGES];
  static bool vnc;
  static bool sdl;
  static bool balloon;
+static bool bin_sh;
  extern bool ioport_debug;
  extern int  active_console;
  extern int  debug_iodelay;
@@ -156,6 +158,8 @@ static const struct option options[] = {
        OPT_BOOLEAN('\0', "balloon",&balloon, "Enable virtio balloon"),
        OPT_BOOLEAN('\0', "vnc",&vnc, "Enable VNC framebuffer"),
        OPT_BOOLEAN('\0', "sdl",&sdl, "Enable SDL framebuffer"),
+       OPT_STRING('\0', "rootfs",&rootfs, "path", "Use path as rootfs"),
+       OPT_BOOLEAN('\0', "binsh",&bin_sh, "Boot rootfs to /bin/sh instead of 
/sbin/init"),

        OPT_GROUP("Kernel options:"),
        OPT_STRING('k', "kernel",&kernel_filename, "kernel",
@@ -615,7 +619,7 @@ int kvm_cmd_run(int argc, const char **argv, const char 
*prefix)
                strlcat(real_cmdline, kernel_cmdline, sizeof(real_cmdline));

        hi = NULL;
-       if (!image_filename[0]) {
+       if (!rootfs&&  !image_filename[0]) {
                hi = host_image(real_cmdline, sizeof(real_cmdline));
                if (hi) {
                        image_filename[0] = hi;
@@ -627,6 +631,18 @@ int kvm_cmd_run(int argc, const char **argv, const char 
*prefix)
        if (!strstr(real_cmdline, "root="))
                strlcat(real_cmdline, " root=/dev/vda rw ", 
sizeof(real_cmdline));

+       if (rootfs) {
+               char tmp[PATH_MAX];
+
+               if (realpath(rootfs, tmp) == 0 ||
+                   virtio_9p__init(kvm, tmp, "/dev/root")<  0)
+                       die("Unable to initialize virtio 9p");
+               strcat(real_cmdline, " root=/dev/root 
rootflags=rw,trans=virtio,version=9p2000.u rootfstype=9p");
+
+               if (bin_sh)
+                       strcat(real_cmdline, " init=/bin/sh");
+       }


Why don't we use the "-d" command line option for this? We can
use fstat() and S_ISDIR to detect that the user specified a directory
as a disk image?

I'm also not completely convinced we should introduce the 'binsh'
command line option. Instead, I'd simply default to 'init=/bin/sh'
unless the user uses -p to specify init.

                        Pekka
--
To unsubscribe from this list: send the line "unsubscribe kvm" in
the body of a message to [email protected]
More majordomo info at  http://vger.kernel.org/majordomo-info.html

Reply via email to