On Sun, 21 May 2000, Andries Brouwer wrote:
> > hda: 0MB, CHS=1108/255/63
> > ^^^^^
> > Partition check:
> > hda: hda1 hda2 hda3
> > ...
> >
> > and kernel cannot mount root fs on hda3.
>
> Yes. That is a bug introduced in 2.3.40 by Al Viro's
> "grok_partitions" patch, if I am not mistaken.
"0Mb" one - guilty as charged, but "cannot mount root" is something else.
> Patch (on 2.3.99p8):
> ----------------------------------------------------------------
> --- hd.c~ Sun Mar 26 13:06:39 2000
> +++ hd.c Sun May 21 01:19:35 2000
> @@ -778,6 +778,8 @@
> #endif
>
> for (drive=0 ; drive < NR_HD ; drive++) {
> + hd[drive<<6].nr_sects = hd_info[drive].head *
> + hd_info[drive].sect * hd_info[drive].cyl;
> printk ("hd%c: %ldMB, CHS=%d/%d/%d\n", drive+'a',
> hd[drive<<6].nr_sects / 2048, hd_info[drive].cyl,
> hd_info[drive].head, hd_info[drive].sect);
> ----------------------------------------------------------------
>
> > Any suggestions how should it be fixed ?
>
> See above. Please complain if this doesnt fix it.
Very odd - check the register_disk() call several lines below. It does exactly
the same assignment and code between these points is not affected by
your change. I would be _very_ surprised if simple substitution
hd_info[drive].sect*...*.... in place of hd[drive<<6].nr_sects would
have an effect different from your patch.
> > Or is the "old disk-only driver" hd.c obsolete and to be dropped ?
>
> The fact that you are the first to complain is an indication
> that not many kernel developers use this old hd.c.
> But it should remain functional.
That's true, but it looks like his problems are not limited to fscked
printk() argument...
-
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@vger.rutgers.edu
Please read the FAQ at http://www.tux.org/lkml/
This archive was generated by hypermail 2b29 : Tue May 23 2000 - 21:00:19 EST