Hi,
I’m trying to install RTP 6.1 on a seperate partition, but always failed with “couldn’t find …qnxbasedma.ifs” error. I tried to use these setups, but all failed :
IDE 0 : quantum 1GB(master), seagate 4.2(slave)
IDE 1 : none(master), NEC 6x CD-ROM(slave)
IDE 0 : quantum1GB(master), no slave
IDE 1 : seagate4.2GB(master), NEC 6x CD-ROM(slave)
IDE 0 : seagate4.2GB(master), NEC 6x CD-ROM(slave)
IDE 1 : nothing connected
Why the setup couldn’t find qnxbasedma.ifs?
TIA
IDE 1 : none(master), NEC 6x CD-ROM(slave)
It’s mentioned in this group over and over and over that the
Masterless IDE is ILLEGAL CONFIGURATION,
no matter “but it works for my favorite OS!”.
kabe
Eduard.
ed1k at yahoo dot com
X Master <xmaster@otakumail.com> wrote in article 1105_1016828100@darkx…
Hi,
I’m trying to install RTP 6.1 on a seperate partition, but always failed with “couldn’t find
…qnxbasedma.ifs” error. I tried to use these setups, but all failed :IDE 0 : quantum 1GB(master), seagate 4.2(slave)
IDE 1 : none(master), NEC 6x CD-ROM(slave)IDE 0 : quantum1GB(master), no slave
IDE 1 : seagate4.2GB(master), NEC 6x CD-ROM(slave)IDE 0 : seagate4.2GB(master), NEC 6x CD-ROM(slave)
IDE 1 : nothing connectedWhy the setup couldn’t find qnxbasedma.ifs?
TIA
Well why shouldn’t it work? My SCSI disks seam to work even though they’re
not properly terminated. ;~}
–
Bill Caroselli – 1(626) 824-7983
Q-TPS Consulting
QTPS@EarthLink.net
<kabe@sra-tohoku.co.jp> wrote in message news:a7iefb$o99$1@inn.qnx.com…
IDE 1 : none(master), NEC 6x CD-ROM(slave)
It’s mentioned in this group over and over and over that the
Masterless IDE is ILLEGAL CONFIGURATION,
no matter “but it works for my favorite OS!”.I dunno why newbies came to believe this config is legal.
kabe
Well why shouldn’t it work? My SCSI disks seam to work even though they’re
not properly terminated. ;~}
I bet you’re just kidding, but facts for others:
PC/AT HDD system was originally an Western Digital WD1003 HDD Controller
sitting on I/O 1F0-1F7,3F6-3F7 IRQ 14, controlling two Seagate ST-506 HDDs.
(Note: not two controller two HDD; ONE controller two HDD,
just like ONE FDD controller (uPD765) and two FDD)
ISA — WD1003 ± ST506
± ST506
Now when IDE was introduced, the WD1003 moved onto the HDD pcboard
and “IDE Interface” became merly ISA bus with an address decoder (#CS0,#CS1)
When you connect TWO IDE disks, TWO former WD1003 controller will
now sit onto the same I/O and IRQ.
That’ll cause electrical/logical conflict.
ISA --(buffer)± [(former WD1003)–HDD]
± [(former WD1003)–HDD]
So you have to tell one of the controller to act as “slave mode” and
only respond upon commands/stats/whatever which only affects the second drive.
Other one will be assigned “master” and responds to commands which
affects the controller itself (common to two HDDs) and things only for
first HDD.
kabe
Hi Kabe
That’s the best explanation of “WHY” you can’t have a slave only disk on
IDE. Hell, even I understood it.
Regarding the unterminated SCSI, yes, I was essentially kidding. But
everyone knows that in a very quiet system (shhhh, the computer is sleeping)
if you only try to do one IO at a time an unterminated SCSI bus will give
the illusion of working.
So you get a few standing waves. Hello, ain’t that what makes PCI so fast?
Maybe we shouldn’t terminate SCSI buses. The terminators probably just slow
all those signals down.
SAFETY NOTICE: DO NOT TRY THIS AT HOME! Bill Caroselli is a known wise ass
and was JUST KIDDING!
<kabe@sra-tohoku.co.jp> wrote in message news:a7q5ef$cbb$1@inn.qnx.com…
Well why shouldn’t it work? My SCSI disks seam to work even though
they’re
not properly terminated. ;~}I bet you’re just kidding, but facts for others:
PC/AT HDD system was originally an Western Digital WD1003 HDD Controller
sitting on I/O 1F0-1F7,3F6-3F7 IRQ 14, controlling two Seagate ST-506
HDDs.
(Note: not two controller two HDD; ONE controller two HDD,
just like ONE FDD controller (uPD765) and two FDD)ISA — WD1003 ± ST506
± ST506Now when IDE was introduced, the WD1003 moved onto the HDD pcboard
and “IDE Interface” became merly ISA bus with an address decoder
(#CS0,#CS1)When you connect TWO IDE disks, TWO former WD1003 controller will
now sit onto the same I/O and IRQ.
That’ll cause electrical/logical conflict.ISA --(buffer)± [(former WD1003)–HDD]
± [(former WD1003)–HDD]So you have to tell one of the controller to act as “slave mode” and
only respond upon commands/stats/whatever which only affects the second
drive.
Other one will be assigned “master” and responds to commands which
affects the controller itself (common to two HDDs) and things only for
first HDD.If you now disconnect the “master” and leave “slave” dangling, IDE
commands
which affects both drives will be unanswered, which will cause
something unpleasant (hang etc)
SCSI bus is designed as a LAN, so every node can work for itself.
Lacking termination will only cause signal degradation and making
cables short will make non-terminated SCSI bus work sometimes.kabe
So you get a few standing waves. Hello, ain’t that what makes PCI so fast?
Maybe we shouldn’t terminate SCSI buses. The terminators probably just slow
all those signals down.
Nope.
PCI is designed “Sustained Tri-state”, which is driven totem-pole.
SCSI is basically Wired-OR, so you NEED pullups somewhere.
(Active terminators will help speeding up the signals)
W-OR will have speed limit, so LVD (Low-voltage Differential) was introduced
for UltraSCSIs.
kabe