lists.openwall.net   lists  /  announce  owl-users  owl-dev  john-users  john-dev  passwdqc-users  yescrypt  popa3d-users  /  oss-security  kernel-hardening  musl  sabotage  tlsify  passwords  /  crypt-dev  xvendor  /  Bugtraq  Full-Disclosure  linux-kernel  linux-netdev  linux-ext4  linux-hardening  linux-cve-announce  PHC 
Open Source and information security mailing list archives
 
Hash Suite: Windows password security audit tool. GUI, reports in PDF.
[<prev] [next>] [<thread-prev] [thread-next>] [day] [month] [year] [list]
Date:	Sun, 5 Nov 2006 18:18:07 +0100 (CET)
From:	Mikulas Patocka <mikulas@...ax.karlin.mff.cuni.cz>
To:	Alan Cox <alan@...rguk.ukuu.org.uk>
Cc:	Albert Cahalan <acahalan@...il.com>, kangur@...com.net,
	linux-kernel@...r.kernel.org
Subject: Re: New filesystem for Linux

>>> And possibly some broken drives may also return you something that
>>> they think is good data but really is not (shouldn't happen since
>>> both disks and cables should be protected by checksums, but hey...
>>> you can never be absolutely sure especially on very big storages).
>
> It happens because
> -	There is limited if any protection on the PCI bus generally
> -	Many PC systems don't have ECC memory, ECC cache
> -	PATA does not CRC protect the command block so if you do enough PATA
> I/O (eg you are a US national lab ..) you *will* eventually get a bit
> flip that gives you the wrong sector with no error data. SATA fixes that
> one.
> -	Murphy is out to get you..

Should IDE driver read back parameters after writing them before issuing 
the command? That should fix this problem. (except when command is written 
badly)

> Not seen that, although they do move stuff aorund in their internal
> block management of bad blocks. I've also seen hardware errors that lead
> to data being messed up silently.

I have seen one WD drive bought in 2003 having error in its firmware in 
cache-coherency code --- if you read and write 256 sectors to the same 
places with some pattern repeatedly (with direct IO), it will discard a 
write. It happens only with 256-sector writes, maybe some part of firmware 
treats 256 as 0. Maybe I create testcase sometimes.

Mikulas

> Alan
>
-
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@...r.kernel.org
More majordomo info at  http://vger.kernel.org/majordomo-info.html
Please read the FAQ at  http://www.tux.org/lkml/

Powered by blists - more mailing lists