Re: i386/pc/grub-setup.c question

2009-08-07 Thread Robert Millan
On Wed, Aug 05, 2009 at 01:41:10AM -0700, Seth Goldberg wrote: > Has the blocklist (--force) code path been tested? We don't test it much. If you test it and find problems, feedback is welcome of course (I think Felix already addressed your concern), but we recommend that you avoid reliing on i

Re: i386/pc/grub-setup.c question

2009-08-05 Thread Seth Goldberg
main () sets the root device if it's not given. grub_file_open doestn't need a device if a root device is set. Else things like `set root=(hd0,1);linux /vmlnuz' wouldn't work either in real grub. It works as long as the path is relative to the root. I.e. grub-install --root-directory=/mnt /dev/s

Re: i386/pc/grub-setup.c question

2009-08-05 Thread Felix Zielcke
Am Mittwoch, den 05.08.2009, 01:41 -0700 schrieb Seth Goldberg: > Hi, > > >I've been looking at grub-setup.c and am perplexed by the following code: > > >/* Make sure that GRUB reads the identical image as the OS. */ >tmp_img = xmalloc (core_size); >core_path_dev = grub_util_ge

i386/pc/grub-setup.c question

2009-08-05 Thread Seth Goldberg
Hi, I've been looking at grub-setup.c and am perplexed by the following code: /* Make sure that GRUB reads the identical image as the OS. */ tmp_img = xmalloc (core_size); core_path_dev = grub_util_get_path (dir, core_file); /* It is a Good Thing to sync two times. */ sync ();