Re: Help - hooking up DSP to XScale running CE6!



No, not a good choice! Virtually every device is mapped as memory or I/O.
There's no reason to try to use something that appears like a "disk" as the
basis for your device, which doesn't look anything like a disk. You want to
define what interface you're going to have, then implement the right driver
under that, not start with the underlying driver and try to figure out how
to shoehorn your DSP into that.

Read the help on stream interface drivers. All they are, as Dean said, is a
DLL with a defined set of exported functions in them. It's dead simple. To
get your driver loaded, you have to add some registry entries, but that's
pretty simple too. Once you have the driver loaded, yes, you can use
ReadFile() and WriteFile() to communicate with the driver, if that makes
sense for what you send to send to and receive from the processor. Or you
can build a set of IOCTL values and use DeviceIoControl() to pass data back
and forth. You might, for example, use WriteFile() to send some raw A/D
data to the processor, then use a DeviceIoControl() message to tell the DSP
to calculate the power spectrum, which you might retrieve via ReadFile().

Paul T.

"Steve Conner" <connermcsteve@xxxxxxxxxxxxxx> wrote in message
news:732dc1ce-0016-4c21-baca-bf1793dc8034@xxxxxxxxxxxxxxxxxxxxxxxxxxxxxx
Hi Dean, thanks for the reply.

You are right, the DSP host port can be mapped into memory.

We have a BSP, but not the source of the BSP. (We're using the
Triton-270 SBC from Ka-Ro and the BSP source was a separate cost
option...) I've already got another application up and running on it,
and I'm now familiar with the process of building OS images, debugging
applications, and so on, but I get really confused when it comes to
device drivers.

I found out that a ramdisk driver ships with CE6, would it be worth
trying that? It seems like I could access my hardware using the same
IO controls you'd use for reading and writing sectors on a disk, and
if I understand right, I'd be more or less doing the same thing when
writing my own driver.

steve


.



Relevant Pages

  • Re: 2.6.23-rc4-mm1 OOPS in forcedeth?
    ... # Generic Driver Options ... # PCI IDE chipsets support ... # SCSI support type (disk, tape, CD-ROM) ...
    (Linux-Kernel)
  • Re: Hardware has finally outrun bloatware?
    ... bottleneck with disk i/o. ... In the case of, say, generations of memory technology ... said that the problem occurred in the display driver. ... The monkey won the money but he scared to pick it up. ...
    (comp.sys.intel)
  • Re: 0x0000007B
    ... If Windows XP Pro cannot install, then you will not get very far with XPe. ... Next I augmented my USB boot with every file system I could find, ... "Windows could not start because of a computer disk hardware configuration ... > Warning:Could not find a driver for Video Controller ...
    (microsoft.public.windowsxp.embedded)
  • Re: RAMDisk and harddisk - Revisited
    ... but your constatation that driver load fails is not accurate in some amount based on info you gave us. ... Only if it is a cloned disk they can have same diskid, but this would be completely different error. ... the ATA Flash Disk was detectable and mountable from RAMDisk image as described. ... e.g. IDE ATA Flash or IDE hard drives. ...
    (microsoft.public.windowsxp.embedded)
  • Re: SCO OS 5.0.7 on Qemu
    ... The OSR5 "wd" IDE driver has code in it to ... > controller used in PC/AT class machines), ... > the kernel to look for a SCSI disk on a "wd" HBA. ... Possible UDMA timing mismatch, ...
    (comp.unix.sco.misc)