[1438] in linux-scsi channel archive
Re: Feb12 release...
daemon@ATHENA.MIT.EDU (Davi Antunes Lima)
Tue Feb 18 00:46:14 1997
Date: Tue, 18 Feb 1997 02:40:27 -0300 (EST)
From: Davi Antunes Lima <davi@tba.com.br>
To: yuri mironoff <yuri@rgti.com>
cc: linux-scsi@vger.rutgers.edu, aic7xxx@FreeBSD.org
In-Reply-To: <Pine.A32.3.95.970216213916.63515B-100000@sce1.rgti.com>
On Sun, 16 Feb 1997, yuri mironoff wrote:
>
>
> On Wed, 12 Feb 1997, Daniel M. Eischen wrote:
>
> >
> > I've placed the Feb12 release of the driver at:
> >
> > ftp://ftp.pcnet.com/users/eischen/Linux/aic7xxx-2.1.23-Feb12.tgz
> >
>
I've got that new release. Thank you for the tip.
> Whats even stranger is that the nice "buffer.c" patch - even though it
> approximately doubles the disk throughput ;))) - accelerates the machine's
> death under bonnie. (ie it keels over immediately after bonnie starts
> running)
I'm currently using an Adaptec 2940 SCSI adapter with a quantum fireball
tm2110s disk. I must confess I'm very disappointed with the disk subsystem
performance. I've already tried anything I think is possible to enhance
those figures, with no success at all. I've already compiled the linux
2.0.29 kernel with the options AIC7XXX_TAGGED_QUEUEING,
AIC7XXX_CMDS_PER_LUN 16, AIC7XXX_PAGE_ENABLE and AIC7XXX_USE_EXT_SCBRAN
enabled, but I've never heard of that "buffer.c" patch.
Can anyone give me a hint on how to improve my SCSI disk throughput by the
use of that "buffer.c" patch or even another one?
Some of you guys from linux-scsi list already told me that the problem
lies on the adaptec 2940 adapter, but I'm not sure of that. To help, I'm
including the kernel msgs, hdparm -t and bonnie results:
aic7xxx: BurstLen = 8 DWDs, Latency Timer = 64 PCLKS
aic7xxx: devconfig = 0x580.
aic7xxx: Reading SEEPROM...done.
aic7xxx: Extended translation enabled.
aic7xxx: Memory check yields 16 SCBs, 255 page-enabled SCBs.
AHA-2940 (PCI-bus), I/O 0xec00, Mem 0xfebff000:
irq 10
bus release time 40 bclks
data fifo threshold 100%
SCSI CHANNEL A:
scsi id 7
scsi selection timeout 256 ms
scsi bus reset at power-on enabled
scsi bus parity enabled
scsi bus termination (low byte) enabled
aic7xxx: Downloading sequencer code...done.
aic7xxx: Resetting the SCSI bus...done.
scsi0 : Adaptec AHA274x/284x/294x (EISA/VLB/PCI-Fast SCSI) 4.0/3.2/4.0
scsi : 1 host.
scsi0: Scanning channel A for devices.
Started kswapd v 1.4.2.2
scsi0: Target 0, channel A, now synchronous at 10.0MHz, offset 15.
Vendor: QUANTUM Model: FIREBALL_TM2110S Rev: 300N
Type: Direct-Access ANSI SCSI revision: 02
Detected scsi disk sda at scsi0, channel 0, id 0, lun 0
scsi0: Enabled tagged queuing for target 0, channel 0, LUN 0, queue depth 8.
scsi : detected 1 SCSI disk total.
SCSI device sda: hdwr sector= 512 bytes. Sectors= 4124736 [2014 MB] [2.0 GB]
Partition check:
sda: sda1 sda2 sda3 sda4
hda: hda1 hda2 hda3 hda4
/dev/sda:
Timing buffered disk reads: 16 MB in 4.76 seconds = 3.36 MB/sec
/dev/sda:
Timing buffer-cache reads: 32 MB in 1.12 seconds =28.57 MB/sec
Bonnie:
-------Sequential Output-------- ---Sequential Input-- --Random--
-Per Char- --Block--- -Rewrite-- -Per Char- --Block--- --Seeks---
Machine MB K/sec %CPU K/sec %CPU K/sec %CPU K/sec %CPU K/sec %CPU /sec %CPU
100 2327 42.3 2297 14.5 1121 14.0 2950 51.9 2271 11.2 29.5 1.8
Thank you for any help!
Davi
