Gabriele Tozzi <gabriele.tozzi <at> gmail.com> writes:

> 
> Hello,
> 
> I got a Promise TX4 SATA 300 Controller with 3 SAMSUNG HD400LJ SATA 300 
> disks (sdc, sdd, sde), the disks runs under a software raid5 with Luks 
> (Cryptography layer) and Lvm2.
> 
> The disks works well and there is no data loss, but i'm continuously and 
> always getting worrisome error messages and poor performance (with brief 
> system jams) when i perform writing operations to disks. Reading data 
> works well and seems to give no error messages.
> 
> I've checked cables and power supply and i've tryied swapping 
> controller's pci slot and playing with bios irq setting. None of them 
> changed the results, so i'm assuming the presence of a kernel bug. I'm 
> sorry if i'm getting wrong and then wasting your time.
> 
> My kernel version:
> Linux version 2.6.21.5 (root <at> transylvania) (gcc version 4.1.3 20070601 
> (prerelease) (Debian 4.1.2-12)) #1 SMP Fri Jun 15 20:25:41 CEST 2007
> 
> This presumed bug was present on linux 2.60.20.7 too.
> 
> On kernel 2.6.18.1 the bug was present, but with different error 
> messages, the messages were identical to those reported by another user 
> on this mail: http://lkml.org/lkml/2006/10/25/198
> 
> I got an Asus CUV4X-D motherboard with two PIII 800Mhz CPUS, this is my 
> home server :)
> 

Everyone seems to be having problems with the Promise SATA300 TX4. I have an 
almost identical setup: SATA300 TX4 with 3 Samsung 400GB drives, and I get very 
similar errors.

In 2.6.20 my card would soft restart all the way down to PIO mode and kill my 
performance. In 2.6.22-rc5 it will hard reset from 3.0 Gb/s to 1.5 Gb/s and 
operate relatively stable until heavy I/O. 

If anyone reading these threads (there have been quite a few in various places, 
see 'sata_promise error handling just a few hours ago) needs any additional 
information to help solve this problem, please let me know.


-
To unsubscribe from this list: send the line "unsubscribe linux-ide" in
the body of a message to [EMAIL PROTECTED]
More majordomo info at  http://vger.kernel.org/majordomo-info.html

Reply via email to