hdparm-1.0 and WD Caviar drives

hdparm-1.0 and WD Caviar drives

Post by Robert Lockhart [Ro » Sat, 27 Aug 1994 03:22:46



I have a Caviar 340MB hard drive with the 128KB cache built-in, and I also
have the hdparm-1.0 IDE utilities.  I've heard that some (particularly WD
Caviar) IDE drives using this feature makes the drive slower than without.
Can anyone give me any feedback on this, or if not, how can I determine for
myself whether it is slowing down or speeding up my system?  I have Linux
kernel 1.1.47 and no IDE cache controller (regular IDE/floppy/IO board).
    -Rob Lockhart

 
 
 

1. hdparm, multiple mode, and WD Caviar 540

Two questions/problems:

1.  MULTIMODE?

I compiled and installed hdparm 1.6 about a week ago, and read in the
man page that enabling multiple mode on some drives ("most notably the
WD Caviar series") actually makes them run slower.

I just got a brand new (only a few weeks old) WD Caviar 540 to replace
my ancient Maxtor SCSI drive which finally died.  Does anyone know
whether enabling multimode will benefit me or not?

I'm using a VL-300 AHA1522-based scsi/ide/serial/parallel controller card
at the moment, so just plain IDE not EIDE.

here's what hdparm reports about the drive:

$ hdparm -i /dev/hda

/dev/hda:

 Model=WDC AC2540H, FwRev=12.08R30WDC AC2540H, SerialNo=WD-WT261
 Config={ HardSect NotMFM HdSw>15uSec SpinMotCtl Fixed DTR>5Mbs FmtGapReq }
 Physical CHS=1048/16/63, TrkSize=57600, SectSize=600, ECCbytes=4
 BuffType=DualPortCache, BuffSize=128KB, MaxMultSect=16
 Features: DblWordIO=no, LBA=yes, DMA=yes, tPIO=1(medium), tDMA=2(fast)
 Current CHS=1048/16/63, TotSect=1056384, MaxLBAsect=1056384
 MultSect=off, DMA-1w=0001, DMA-mw=0003
 EIDE: IORDY=switchable, mode3, tDMA={min:150,rec:150}, tPIO={min:330,iordy:180}

$ hdparm /dev/hda
/dev/hda:
 readahead = 8
 multcount = 0
 readonly  = 0
 unmaskirq = 0
 geometry  = 1048/16/63, offset = 0
 sectors   = 1056384

2.  hdparm -t crashes?

When I first installed hdparm, I enabled 16 sector multimode, then ran
"hdparm -t" whilst compiling a new version of the kernel in another VC.
I didn't expect any meaningful results (because of the disk thrashing
caused by the compile) but I also certainly didn't expect the system to
lock up.  All disk activity ceased, but I was still able to switch
consoles.  Couldn't do anything so eventually had to press the reset
button.

When I get time I'll reboot my machine into single user mode and test it
more thoroughly.  In the meantime, has anyone else noticed anything like
this?

--

    "Organise a strike in your school or workplace on the grounds that it
     does not satisfy your need for indolence & spiritual beauty."
                              -- Hakim Bey, The Temporary Autonomous Zone

2. Setting stty permanently

3. hdparm WD Caviar 1.2 gig and 1.2.9

4. Unix - Frequently Asked Questions (Contents) [Frequent posting]

5. HARD DRIVE WD CAVIAR AC2700 (LILO)

6. slice program

7. LBA/biosgeom, WD Caviar 1.2Gb drive

8. Sanity check on my network setup

9. Summary, WD 250meg Caviar Drive

10. WD Caviar 1.2 GB drives

11. WD Caviar Drives

12. trouble with WD 1.6gb Caviar drive

13. WD-Caviar-drive is too slow under Linux.