Installing/booting Visopsys from USB?

[El_Matador]

1) I have tried both options
2) hd0a and cd0
3) no errors

I think it isn't a Visopsys related problem. Some other OSes don't recognize it.

Is it possible to use unetbootin or a similar tool to create a booteable USB drive for Visopsys?

[andymc]

It could be an issue with the stick or the port. My development system has a USB port on the front panel that never works with Visopsys, and often causes trouble for Linux too. The ports on the back of the machine tend to work. I understand that it might have something to do with power (front port doesn't have the same level of available power, or something).

It might be a slightly fishy stick, too, and/or something that Visopsys isn't doing correctly. At the present time, Visopsys works with all dozen-or-so sticks that I own, running on real hardware. I've noticed some of them don't seem to work as well in VmWare since my last round of USB work but that's the way it tends to go with hardware devices: It's tough to get everything just right for every system and every device.

The best way to make a bootable Visopsys USB right now would be to use the installer, as you already tried. When I do the next release (which is actually, basically ready - I've been working on a compression library for a future release) you'll be able to use scripts in the utils/ directory to create USB images.

Here's a USB image that I just created for you:
http://visopsys.org/files/misc/visopsys ... sb-img.zip

You can use a program like this one to write it to a stick:
https://launchpad.net/win32-image-writer/+download

If you're working with source code, I could give you the scripts to create USB images, but the last version of Visopsys really won't boot from USB anyway. The new release should be available in the next few days anyway.

Andy

[El_Matador]

It could be an issue with the stick or the port.

The issue was with the stick.

you'll be able to use scripts in the utils/ directory to create USB images.

I will be using them.

Here's a USB image that I just created for you:
http://visopsys.org/files/misc/visopsys ... sb-img.zip

Thanks that works

You can use a program like this one to write it to a stick:
https://launchpad.net/win32-image-writer/+download

Writing to a physical device can corrupt the device.
Are you sure you want to continue?
-Yes

Now the stick has a total of 29.9MB but Visopsys now runs on my real hardware
PS: I wanted to boot Visopsys from USB because the computer I'm using is a netbook (and it doesn't have a CD-ROM drive)

[andymc]

Writing to a physical device can corrupt the device.
Are you sure you want to continue?
-Yes

that message tries to discourage people from writing directly to devices, unless they understand what they're doing (for example, clobbering the partition table)

Now the stick has a total of 29.9MB but Visopsys now runs on my real hardware
PS: I wanted to boot Visopsys from USB because the computer I'm using is a netbook (and it doesn't have a CD-ROM drive)

Yes, the image has a partition table in its MBR sector. I had to pick a partition size, so I picked one that will fit on even the smallest/oldest USB sticks, but still big enough for Visopsys. If you reformat the stick in Windows, or resize the partition, you'll get all your old space back.

Andy

[El_Matador]

What does this error mean? (I'm booting Visopsys from USB)

[andymc]

What does this error mean? (I'm booting Visopsys from USB)

Hiya,

The first one is a non-fatal error that means the USB (SCSI) disk driver wasn't able to guess disk geometry numbers that match up with the number of sectors on the disk. USB disks don't really have geometry, but certain things (notably the disk partitioning code) expect there to be a 'cylinders' value, for example. The code that gave that error wasn't able to guess something that made sense.

The second one is usually seen when an IDE disk claims to support a DMA/UDMA mode, but an attempt to read sectors using that mode failed. The IDE code will try successively lower modes until it finds one that works, but it does produce those error messages when a mode fails.

The third message looks like a minor bug, an attempt to re-deallocate memory, possibly caused by the removal of a USB stick that wasn't enumerated correctly?

[El_Matador]

The first one is a non-fatal error that means the USB (SCSI) disk driver wasn't able to guess disk geometry numbers that match up with the number of sectors on the disk. USB disks don't really have geometry, but certain things (notably the disk partitioning code) expect there to be a 'cylinders' value, for example. The code that gave that error wasn't able to guess something that made sense.

The second one is usually seen when an IDE disk claims to support a DMA/UDMA mode, but an attempt to read sectors using that mode failed. The IDE code will try successively lower modes until it finds one that works, but it does produce those error messages when a mode fails.

Thanks for the clarification.

The third message looks like a minor bug, an attempt to re-deallocate memory, possibly caused by the removal of a USB stick that wasn't enumerated correctly?

I haven't removed my USB stick. What else can be causing that problem?

[andymc]

The third message looks like a minor bug, an attempt to re-deallocate memory, possibly caused by the removal of a USB stick that wasn't enumerated correctly?

I haven't removed my USB stick. What else can be causing that problem?

Hmm, not sure. If I've interpreted this correctly, it's not really a 'problem' per se -- the system caught the error -- it just turned up a minor coding error on my part; the hotplug device detection frees the memory it allocated if there's an error along the way, but doesn't set the pointers to NULL. The hotplug device removal code sees the non-NULL pointers and tries to free the memory again.