Re: hurd or Mach hangs at regular intervals

[Rahul Saxena]

Hi, i am attaching 2 msgs from the email archive i found on the mailing 
lists, for reference. it seems somone had this 'bad access' error earlier 
too. 
my original msg thread follows after that.

has it been resolved since ? 

rahul

********
http://www.geocrawler.com/mail/msg.php3?msg_id=2377359&list=332
********
FROM: OKUJI Yoshinori
DATE: 06/28/1999 01:28:51
SUBJECT: RE:  hd error during bootstrap

From: "Momchil `Velco` Velikov" <<EMAIL: PROTECTED>>
Subject: hd error during bootstrap
Date: Sun, 27 Jun 1999 15:29:24 +0300

> hd01: bad access : block=28, count=2, blockend=30, nr_sects=2
> end_request: I/O error, dev 03:01, sector  28
> Hit boot to return...
> (serverboot): Running without any paging

  Thanks for the clue. This seems to be due to a bug in Hurd but not
in Mach. The Linux partition handling code (genhd.c in gnumach) sets
NR_SECTS to 2 if NR_SECTS of an extended partition is more than 2,
because normal programs should not use any extended partition. An
extended partition should be used only to put logical partitions into
it. Therefore, I suspect that serverboot (and/or other program in
Hurd) tries to read 2 sectors starting at 28 from any partition,
whether it is an extended partition or not.
**********
**********
http://www.geocrawler.com/mail/msg.php3?msg_id=2378927&list=332
**********
FROM: Thomas Bushnell, BSG
DATE: 06/28/1999 11:06:06
SUBJECT: RE:  hd error during bootstrap

OKUJI Yoshinori <<EMAIL: PROTECTED>> writes:

>   Thanks for the clue. This seems to be due to a bug in Hurd but not
> in Mach. The Linux partition handling code (genhd.c in gnumach) sets
> NR_SECTS to 2 if NR_SECTS of an extended partition is more than 2,
> because normal programs should not use any extended partition. An
> extended partition should be used only to put logical partitions into
> it. Therefore, I suspect that serverboot (and/or other program in
> Hurd) tries to read 2 sectors starting at 28 from any partition,
> whether it is an extended partition or not.

I don`t think it`s a bug in the Hurd; it`s a bug in whoever decides to
page on a partition that isn`t big enough.  
**********

 
On Thursday 11 September 2003 07:50 pm, Rahul Saxena wrote:
> On Thursday 11 September 2003 02:23 pm, Alfred M. Szmidt wrote:
> >    BTW i have also been getting ext2fs.static:hd02:BAD ACCESS errors
> >    for blocks starting from 28 to 30 on my harddrive.
> >
> > Erm, what is hd02? Thats not a valid device name in GNU/Hurd... It
> > should be hd2, hd2s1 or some such.
>
> yes but thats what i get... i have got this error a few times and its hd02,
> i know that it should be hd2 or with some slice number but this is what i
> get.
>
> > And next time, please tell us the _exact_ error message, the string
> > "BAD ACCESS" does not even exist in GNU Mach, or the Hurd (you
> > probobly mean "bad access" which is in
> > gnumach/linux/dev/drivers/block/ide.c)
>
> yes sorry about that. thats bad access only. i had written the errors down.

-- 
A bird in the hand is worth what it will bring.