VIA vt82c686b and UDMA(100)

VIA vt82c686b and UDMA(100)

[Brian Dushaw]

Dear Linux Kernel Wisemen,
   I have been following the discussion of the VIA vt82c686b chipset
and the troubles people have had in getting UDMA(100) to work.  This
is to report that I have now tried the 2.4.2-ac20 kernel and the
2.2.18 kernel with Andre's patch (dated March 20) and neither of
them get the disk speed up to where it ought to be.  hdparm -t reports
back 11 MB/s or so for either kernel.
   VIA82CXXX enabled, and I also tried the ide0=ata66 flag, in desparation.
   At boot up both kernels report the disk as UDMA(100) - everything
seems to be peachy keen, but for the sluggish disk performance.

Merely a report from the front lines,

B.D.

Re: VIA vt82c686b and UDMA(100)

[William Park]

On Wed, Mar 21, 2001 at 08:40:21PM -0800, Brian Dushaw wrote:
> Dear Linux Kernel Wisemen,
>    I have been following the discussion of the VIA vt82c686b chipset
> and the troubles people have had in getting UDMA(100) to work.  This
> is to report that I have now tried the 2.4.2-ac20 kernel and the
> 2.2.18 kernel with Andre's patch (dated March 20) and neither of
> them get the disk speed up to where it ought to be.  hdparm -t reports
> back 11 MB/s or so for either kernel.
>    VIA82CXXX enabled, and I also tried the ide0=ata66 flag, in desparation.
>    At boot up both kernels report the disk as UDMA(100) - everything
> seems to be peachy keen, but for the sluggish disk performance.
> 
> Merely a report from the front lines,

Try 'hdparm -d1 -t', and see what you get.

:wq --William Park, Open Geometry Consulting, Linux/Python, 8 CPUs.

Re: VIA vt82c686b and UDMA(100)

[Brian Dushaw]

No luck with either William's or Agus's suggestions.  Still 11 MB/s
transfer rate, dma enabled or not.  The motherboard is a newer IWILL.

dmesg outputs:
Uniform Multi-Platform E-IDE driver Revision: 6.30
ide: Assuming 33MHz system bus speed for PIO modes; override with idebus=xx
VP_IDE: IDE controller on PCI bus 00 dev 39
VP_IDE: chipset revision 6
VP_IDE: not 100% native mode: will probe irqs later
ide: Assuming 33MHz system bus speed for PIO modes; override with idebus=xx
VP_IDE: VIA vt82c686b (rev 40) IDE UDMA100 controller on pci00:07.1
    ide0: BM-DMA at 0xd000-0xd007, BIOS settings: hda:DMA, hdb:pio
    ide1: BM-DMA at 0xd008-0xd00f, BIOS settings: hdc:DMA, hdd:pio
hda: WDC WD300BB-00AUA1, ATA DISK drive
hdc: Hewlett-Packard CD-Writer Plus 9100b, ATAPI CDROM drive
ide0 at 0x1f0-0x1f7,0x3f6 on irq 14
ide1 at 0x170-0x177,0x376 on irq 15
hda: WDC WD300BB-00AUA1, 28629MB w/2048kB Cache, CHS=3649/255/63, UDMA(100)

hdparm -i /dev/hda outputs:
 Model=WDC WD300BB-00AUA1, FwRev=18.20D18, SerialNo=WD-WMA6R3063544
 Config={ HardSect NotMFM HdSw>15uSec SpinMotCtl Fixed DTR>5Mbs FmtGapReq }
 RawCHS=16383/16/63, TrkSize=57600, SectSize=600, ECCbytes=40
 BuffType=3(DualPortCache), BuffSize=2048kB, MaxMultSect=16, MultSect=off
 DblWordIO=no, OldPIO=2, DMA=yes, OldDMA=0
 CurCHS=16383/16/63, CurSects=-66060037, LBA=yes, LBAsects=58633344
 tDMA={min:120,rec:120}, DMA modes: mword0 mword1 mword2
 IORDY=on/off, tPIO={min:120,w/IORDY:120}, PIO modes: mode3 mode4
 UDMA modes: mode0 mode1 mode2 mode3 mode4 *mode5

hdparm -t /dev/hda outputs:
/dev/hda:
 Timing buffered disk reads:  64 MB in  5.66 seconds = 11.31 MB/sec

[sigh...]

I suppose it could be the Western Digital disk - I seem to recall that
linux has a difficult history with WD (a comment that may merely start
an unfounded rumour...)

B.D.

On Thu, 22 Mar 2001, William Park wrote:

> On Wed, Mar 21, 2001 at 08:40:21PM -0800, Brian Dushaw wrote:
> > Dear Linux Kernel Wisemen,
> >    I have been following the discussion of the VIA vt82c686b chipset
> > and the troubles people have had in getting UDMA(100) to work.  This
> > is to report that I have now tried the 2.4.2-ac20 kernel and the
> > 2.2.18 kernel with Andre's patch (dated March 20) and neither of
> > them get the disk speed up to where it ought to be.  hdparm -t reports
> > back 11 MB/s or so for either kernel.
> >    VIA82CXXX enabled, and I also tried the ide0=ata66 flag, in desparation.
> >    At boot up both kernels report the disk as UDMA(100) - everything
> > seems to be peachy keen, but for the sluggish disk performance.
> >
> > Merely a report from the front lines,
>
> Try 'hdparm -d1 -t', and see what you get.
>
> :wq --William Park, Open Geometry Consulting, Linux/Python, 8 CPUs.
>

-- 
%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%

Brian Dushaw
Applied Physics Laboratory
University of Washington
1013 N.E. 40th Street
Seattle, WA  98105-6698
(206) 685-4198   (206) 543-1300
(206) 543-6785 (fax)
dushaw@apl.washington.edu

Web Page:  http://staff.washington.edu/dushaw/index.html

Re: VIA vt82c686b and UDMA(100)

[Brian Dushaw]

And for the record:

"hdparm -d1 -t -X69 /dev/hda" gives:

/dev/hda:
 setting using_dma to 1 (on)
 setting xfermode to 69 (UltraDMA mode5)
 using_dma    =  1 (on)
 Timing buffered disk reads:  64 MB in  5.64 seconds = 11.35 MB/sec

-- 
%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%

Brian Dushaw
Applied Physics Laboratory
University of Washington
1013 N.E. 40th Street
Seattle, WA  98105-6698
(206) 685-4198   (206) 543-1300
(206) 543-6785 (fax)
dushaw@apl.washington.edu

Web Page:  http://staff.washington.edu/dushaw/index.html

Re: VIA vt82c686b and UDMA(100)

[Nils Philippsen]

On Thu, 22 Mar 2001, Brian Dushaw wrote:

> And for the record:
>
> "hdparm -d1 -t -X69 /dev/hda" gives:

My current hdparm line looks like this:

hdparm -m16 -c1 -u1 -k1 -X69 /dev/...

With this, I can get 28.x MB/s instead of 15.y with just -X69.

Nils
-- 
 Nils Philippsen / Berliner Straße 39 / D-71229 Leonberg // +49.7152.209647
nils@wombat.dialup.fht-esslingen.de / nils@fht-esslingen.de / nils@redhat.de
   The use of COBOL cripples the mind; its teaching should, therefore, be
   regarded as a criminal offence.                  -- Edsger W. Dijkstra

Re: VIA vt82c686b and UDMA(100)

[Brian Dushaw]

In response to some suggestions I give
more for the record on my problem of getting
ata100 transfer rates:

"hdparm /dev/hda" gives:

/dev/hda:
 multcount    =  0 (off)
 I/O support  =  1 (32-bit)
 unmaskirq    =  1 (on)
 using_dma    =  1 (on)
 keepsettings =  0 (off)
 nowerr       =  0 (off)
 readonly     =  0 (off)
 readahead    =  8 (on)
 geometry     = 3649/255/63, sectors = 58633344, start = 0

so 32 bit mode and dma are indeed on... and still 11 MB/s.

Thx for all the suggestions!
B.D.

-- 
%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%

Brian Dushaw
Applied Physics Laboratory
University of Washington
1013 N.E. 40th Street
Seattle, WA  98105-6698
(206) 685-4198   (206) 543-1300
(206) 543-6785 (fax)
dushaw@apl.washington.edu

Web Page:  http://staff.washington.edu/dushaw/index.html

Re: VIA vt82c686b and UDMA(100)

[David Balazic]

Brian Dushaw (dushaw@munk.apl.washington.edu) wrote :

> Dear Linux Kernel Wisemen, 
>    I have been following the discussion of the VIA vt82c686b chipset 
> and the troubles people have had in getting UDMA(100) to work. This 
> is to report that I have now tried the 2.4.2-ac20 kernel and the 
> 2.2.18 kernel with Andre's patch (dated March 20) and neither of 
> them get the disk speed up to where it ought to be. hdparm -t reports 
> back 11 MB/s or so for either kernel. 
>    VIA82CXXX enabled, and I also tried the ide0=ata66 flag, in desparation. 
>    At boot up both kernels report the disk as UDMA(100) - everything 
> seems to be peachy keen, but for the sluggish disk performance. 
> 
> Merely a report from the front lines, 
> 
> B.D. 

Do you also have IDE_AUTO_WHATEVER option enabled , 
as suggested ( no, commanded ) in the VIA_IDE_OPTION help text ?
( press '?' when selecting the VIA IDE driver option )

cat /proc/ide/via ?

What do you think is the "correct" transfer rate of the disk ?

For the record , I have a MSI K7T Pro2A board ( VIA KT133 with a
vt82c686b south bridge ) and a IBM DTLA 307045 hard drive on a 80
wire IDE cable ( set to CABLE-SELECT , connected to the end connector;
you must always first use both connectors on the end of the cable !
never left one end unused )

Without doing any settings with hdparm, I get the full transfer rate
of the disk, measured with hdparm : ~35MB/s


kernel is 2.4.recent or redhat recent 2.4.x versions.

-- 
David Balazic
--------------
"Be excellent to each other." - Bill & Ted
- - - - - - - - - - - - - - - - - - - - - -

Re: VIA vt82c686b and UDMA(100)

[Brian Dushaw]

Thanks to all for their advice on this problem.  So far I've tried
hdparm flags six ways from sunday to no effect, but I have yet to
try suggestions such as those below.  Alas I have to set aside this
problem for the next few weeks.

In the mean time, I note that the linux machines at my Lab are being
subjected to fairly severe security attacks at the moment (the latest
involved a hold in the bind DNS server (?) ) - a reminder
to all to keep applying those updates and security patches.

B.D.

On Fri, 23 Mar 2001, David Balazic wrote:

> Brian Dushaw (dushaw@munk.apl.washington.edu) wrote :
>
> > Dear Linux Kernel Wisemen,
> >    I have been following the discussion of the VIA vt82c686b chipset
> > and the troubles people have had in getting UDMA(100) to work. This
> > is to report that I have now tried the 2.4.2-ac20 kernel and the
> > 2.2.18 kernel with Andre's patch (dated March 20) and neither of
> > them get the disk speed up to where it ought to be. hdparm -t reports
> > back 11 MB/s or so for either kernel.
> >    VIA82CXXX enabled, and I also tried the ide0=ata66 flag, in desparation.
> >    At boot up both kernels report the disk as UDMA(100) - everything
> > seems to be peachy keen, but for the sluggish disk performance.
> >
> > Merely a report from the front lines,
> >
> > B.D.
>
> Do you also have IDE_AUTO_WHATEVER option enabled ,
> as suggested ( no, commanded ) in the VIA_IDE_OPTION help text ?
> ( press '?' when selecting the VIA IDE driver option )
>
> cat /proc/ide/via ?
>
> What do you think is the "correct" transfer rate of the disk ?
>
> For the record , I have a MSI K7T Pro2A board ( VIA KT133 with a
> vt82c686b south bridge ) and a IBM DTLA 307045 hard drive on a 80
> wire IDE cable ( set to CABLE-SELECT , connected to the end connector;
> you must always first use both connectors on the end of the cable !
> never left one end unused )
>
> Without doing any settings with hdparm, I get the full transfer rate
> of the disk, measured with hdparm : ~35MB/s
>
>
> kernel is 2.4.recent or redhat recent 2.4.x versions.
>
>

-- 
%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%

Brian Dushaw
Applied Physics Laboratory
University of Washington
1013 N.E. 40th Street
Seattle, WA  98105-6698
(206) 685-4198   (206) 543-1300
(206) 543-6785 (fax)
dushaw@apl.washington.edu

Web Page:  http://staff.washington.edu/dushaw/index.html