Bug#336519: bad initrd_size?

[dann frazier <dannf@hp.com>]

hey Vagrant,
  I just root caused an issue w/ the same symptom on ia64.  Turns out
the bootloader was passing a bloated initrd_size option to the kernel.
Since initramfs is known to work on x86 w/ other bootloaders, I'm
thinking qemu maybe doing the same thing.

Can you try a test for me?

1) Rebuild your kernel w/ the following patch, and post a copy of the
boot log
2) Post the output of zcat /boot/initrd-img-2.6.14-1-386 | wc -c

diff --git a/init/initramfs.c b/init/initramfs.c
--- a/init/initramfs.c
+++ b/init/initramfs.c
@@ -416,6 +416,10 @@ static void __init flush_window(void)
 static char * __init unpack_to_rootfs(char *buf, unsigned len, int check_only)
 {
        int written;
+       int firstok = 0;
+
+       printk(KERN_INFO "DANNF: initramfs.c:unpack_to_rootfs(%p, %d, %d)\n",
+              buf, len, check_only);
        dry_run = check_only;
        header_buf = malloc(110);
        symlink_buf = malloc(PATH_MAX + N_ALIGN(PATH_MAX) + 1);