lkml.org 
[lkml]   [1998]   [May]   [29]   [last100]   RSS Feed
Views: [wrap][no wrap]   [headers]  [forward] 
 
Messages in this thread
/
Date
From
SubjectRe: Confirmation; Severe ext2fs Corruption in 2.1.98
On Fri, 29 May 1998, Phil's Kernel Account wrote:
>
> Now, think about this. Here's the file...
>
> -rw-r--r-- 1 root root 3968254 May 6 23:31
> Robert_Pohl-Master_Of_Evil.mp3
>
> It was being moved from /home/irc (hdb2) to /mp3/dance (sda1). As you can
> see, this file would most likely be buffered in it's entirety. Now

I'll throw in my two sense (sic) here. I use SCSI drives. I had this
problem several times. However, the problem didn't destroy everything. I
found that a bad inode entry, that attempts to access beyond the physical
media, does not seem to be limit-checked anywhere. Therefore, with SCSI,
the driver will fail <forever> as it keeps getting errors from the drive,
keeps resetting the bus, card, driver, etc., then tries to access beyond
the drive again and again and again. Boot-switch time.

Now, with IDE, since the head/cylinder/sector translation is virtual and
cheap, they will wrap the drive. In other words, warp past some max and
back up through zero. The result is the access doesn't fail! Instead,
you eventually write something where it shouldn't be.

I think that some code somewhere should really check access limits.

Cheers,
Dick Johnson
***** FILE SYSTEM MODIFIED *****
Penguin : Linux version 2.1.103 on an i586 machine (66.15 BogoMips).
Warning : It's hard to remain at the trailing edge of technology.


-
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@vger.rutgers.edu

\
 
 \ /
  Last update: 2005-03-22 13:42    [W:0.047 / U:0.376 seconds]
©2003-2020 Jasper Spaans|hosted at Digital Ocean and TransIP|Read the blog|Advertise on this site