I wonder if I am doing something wrong or what. It looks like the IDE driver
devb-eide in Neutrino 2.0 (Patch B) is taking 20 seconds to initialize. I
have one hard disk on the primary channel and another one on the secondary
channel. I am automounting the second disk as /
Is this 20 seconds delay normal?
-Kim
Personally I think not, but dev-eide tries to auto-detect everything.
I was told that if you provide it with all the proper argument (
so it doesn’t autodetect) it will start much faster.
“Kim Liu” <kliu@terayon.com> wrote in message
news:snfa98nb6tt183@corp.supernews.com…
I wonder if I am doing something wrong or what. It looks like the IDE
driver
devb-eide in Neutrino 2.0 (Patch B) is taking 20 seconds to initialize. I
have one hard disk on the primary channel and another one on the secondary
channel. I am automounting the second disk as /Is this 20 seconds delay normal?
-Kim
Sometimes, yes. Depends on model of drives and controller.
Kim Liu wrote:
I wonder if I am doing something wrong or what. It looks like the IDE driver
devb-eide in Neutrino 2.0 (Patch B) is taking 20 seconds to initialize. I
have one hard disk on the primary channel and another one on the secondary
channel. I am automounting the second disk as /Is this 20 seconds delay normal?
-Kim
Sometimes, yes. Depends on model of drives and controller.
Well, what I have is a regular Pentium II PC with the 440BX chipset. I have
a regular Quantum 4.2GB hard disk as the primary master device and a
M-Systems 8MB flash device with a ATA interface as the secondary master
device. The motherboard BIOS is set up to boot from the flash device.
In any case, why does it take so long to initialize anyway? I don’t think
the EIDE driver in my QNX 4.25 takes more than 2 seconds.
So I have been experimenting with some of the eide options (without fully
understanding what they do) hoping that the EIDE driver can initialize
faster. Here’s my finding so far:
No option, everything auto: 17 sec
With the “slave” option (don’t detect slave devices): 10 sec
The other options (nopci, cmd640, timeout, nobios) don’t seem to matter.
Then I detached the hard disk (on the primary channel) and moved the ATA
flash device to the primary channel, hoping that one less device will make
it faster. Turned out it still took 17 seconds but what’s strange is that
the EIDE drover detected bogus master and slave device on the secondary
channel (which is not physically connected to anything).
Finally, I disabled the secondary channel completely from the BIOS. With
this setting, the EIDE driver finished its job very quickly (like 1 second).
Can someone explain to me what’s going on?
-Kim
I think the problem may be with devb-eide desperately trying to identify
devices on secondary channel. It has to be rather persistent in that,
because some of IDE devices are really nasty about replying to ATA
IDENTIFY. You can turn off secondary interface in the driver, this
should make it faster.
Kim Liu wrote:
Sometimes, yes. Depends on model of drives and controller.
Well, what I have is a regular Pentium II PC with the 440BX chipset. I have
a regular Quantum 4.2GB hard disk as the primary master device and a
M-Systems 8MB flash device with a ATA interface as the secondary master
device. The motherboard BIOS is set up to boot from the flash device.In any case, why does it take so long to initialize anyway? I don’t think
the EIDE driver in my QNX 4.25 takes more than 2 seconds.So I have been experimenting with some of the eide options (without fully
understanding what they do) hoping that the EIDE driver can initialize
faster. Here’s my finding so far:No option, everything auto: 17 sec
With the “slave” option (don’t detect slave devices): 10 sec
The other options (nopci, cmd640, timeout, nobios) don’t seem to matter.Then I detached the hard disk (on the primary channel) and moved the ATA
flash device to the primary channel, hoping that one less device will make
it faster. Turned out it still took 17 seconds but what’s strange is that
the EIDE drover detected bogus master and slave device on the secondary
channel (which is not physically connected to anything).Finally, I disabled the secondary channel completely from the BIOS. With
this setting, the EIDE driver finished its job very quickly (like 1 second).Can someone explain to me what’s going on?
-Kim
Click here for Free Video!!
http://www.gohip.com/free_video/