On Friday 28 December 2007 19:56:24 Balaji Rao wrote:
> hi rusty,
>
> here's the patch with the mistakes corrected.
Applied. It didn't compile though: devices.device should be device.dev.
I tested it here, and looking at /proc/<pid>/fd, some fds were still leaking
(for example, the /dev/lguest fd, which means we were keeping guests around in
the kernel, and also we never closed the vmlinuz fd).
So I decided your original approach was correct, sorry. Here's the patch I
applied on top. Note the 80 column limit, and spaces in "if (" but not
"close(".
Thanks!
Rusty.
diff -r c0e1cb9a4bba Documentation/lguest/lguest.c
--- a/Documentation/lguest/lguest.c Sat Jan 05 09:58:14 2008 +1100
+++ b/Documentation/lguest/lguest.c Sat Jan 05 10:46:31 2008 +1100
@@ -1527,8 +1527,8 @@ static void setup_block_file(const char
/* Create stack for thread and run it */
stack = malloc(32768);
- /* SIGCHLD - We dont "wait" for our cloned thread, so prevent it from
becoming
- * a zombie. */
+ /* SIGCHLD - We dont "wait" for our cloned thread, so prevent it from
+ * becoming a zombie. */
if (clone(io_thread, stack + 32768, CLONE_VM | SIGCHLD, dev) == -1)
err(1, "Creating clone");
@@ -1587,26 +1587,22 @@ static void setup_rng(void)
verbose("device %u: rng\n", devices.device_num);
}
-/* That's the end of device setup. */
+/* That's the end of device setup. :*/
-/* Restart the guest */
-static void restart_guest(void)
+/* Reboot */
+static void __attribute__((noreturn)) restart_guest(void)
{
- struct device *dev;
- struct vblk_info *vblk;
+ unsigned int i;
- /* Closing the waked_fd causes the waker thread to die */
- close (waker_fd);
- /* Closing the workpipe[1] causes io_thread thread to die */
- for(dev = devices.device;dev!= NULL;dev = dev->next) {
- if (strcmp(dev->name, "block") == 0) {
- vblk = dev->priv;
- close (vblk->workpipe[1]);
- }
- }
- if(execv(main_args[0],main_args))
- errx(1,"Could not exec %s", main_args[0]);
+ /* Closing pipes causes the waker thread and io_threads to die, and
+ * closing /dev/lguest cleans up the Guest. Since we don't track all
+ * open fds, we simply close everything beyond stderr. */
+ for (i = 3; i < FD_SETSIZE; i++)
+ close(i);
+ execv(main_args[0], main_args);
+ err(1, "Could not exec %s", main_args[0]);
}
+
/*L:220 Finally we reach the core of the Launcher, which runs the Guest, serves
* its input and output, and finally, lays it to rest. */
static void __attribute__((noreturn)) run_guest(int lguest_fd)
@@ -1681,6 +1677,7 @@ int main(int argc, char *argv[])
/* If they specify an initrd file to load. */
const char *initrd_name = NULL;
+ /* Save the args: we "reboot" by execing ourselves again. */
main_args = argv;
/* We don't "wait" for the children, so prevent them from becoming
* zombies. */
_______________________________________________
Lguest mailing list
[email protected]
https://ozlabs.org/mailman/listinfo/lguest
