Pci Device Driver Qnx - programmes-guru’s blog

Typically drivers use this function to attach themselves to a PCI device, so that other drivers can t attach to the same device. If you specify the PCI_SHARE flag.

PCI Drivers Download Utility at Drivers.com - Update PCI Drivers for your PC - Free Drivers Scan Automatic Updates.

Pci_find_device Find the PCI device with a given device ID and vendor ID. Synopsis: include hw/pci.h int pci_find_device unsigned device, unsigned.

QNX realtime RTOS - Operating systems, Detach a driver from a PCI device. Synopsis: include hw/pci.h int pci_detach_device void handle ; Arguments: handle.

Connecting Hardware

This chapter includes:

Introduction

When you boot a Neutrino desktop system, it starts a

device enumerator,

a manager that detects most hardware devices.

The enumerator loads a set of configuration files from

/etc/system/enum/devices that

define what your system should do e.g. start a specific driver

when you add or remove hardware.

You can edit the enumerator s configuration files, if necessary.

For more information, see

Controlling How Neutrino Starts

in this guide, and

enum-devices

in the Utilities Reference.

An embedded Neutrino system typically has specific hardware, so when the

system boots, it s likely to explicitly start the appropriate drivers.

You can find a list of currently supported hardware in the

Developer Support Center area of our website,

The website lists the chipsets and hardware that we ve tested with Neutrino.

However, many times there are slight variants of chipsets that

will work with the drivers even if they aren t listed.

It s often worth trying these chipsets to see if the driver

will work with your hardware, but note that the hardware might not

behave as expected.

Neutrino doesn t currently support tapes.

You ll use the information in this chapter if the enumerator can t detect

your system s devices, or if you want to manually configure static devices

in an embedded system.

You need to be logged in as root to start any drivers.

Make sure that PnP-aware OS is disabled in the BIOS before you

run Neutrino.

PCI/AGP devices

If you don t know what type of controller you re using, you can

use the

pci

utility to identify it:

pci -vvv less

The output from this command looks something like this:

Class Mass Storage IDE

Vendor ID 8086h, Intel Corporation

Device ID 7111h, 82371AB/EB PIIX4 IDE Controller

PCI index 0h

Class Codes 010180h

Revision ID 1h

Bus number 0

Device number 4

Function num 1

Status Reg 280h

Command Reg 5h

I/O space access enabled

Memory space access disabled

Bus Master enabled

Special Cycle operations ignored

Memory Write and Invalidate disabled

Palette Snooping disabled

Parity Checking disabled

Data/Address stepping disabled

SERR driver disabled

Fast back-to-back transactions to different agents disabled

Header type 0h Single-function

BIST 0h Build-in-self-test not supported

Latency Timer 20h

Cache Line Size 0h

PCI IO Address d800h length 16 enabled

Max Lat 0ns

Min Gnt 0ns

PCI Int Pin NC

Interrupt line 0

Device Dependent Registers:

0x40: 07 c0 03 80 00 00 00 00 05 00 02 02 00 00 00 00

0x50: 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00

0x60: 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00

0x70: 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00

0x80: 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00

0x90: 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00

0xA0: 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00

0xB0: 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00

0xC0: 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00

0xD0: 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00

0xE0: 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00

0xF0: 00 00 00 00 00 00 00 00 30 0f 00 00 00 00 00 00

Find the entry for the device you want to locate and it ll give

you the details on the manufacturer/vendor ID and device ID. You

may need to search for keywords e.g. Audio in order to identify

your device.

You can search the manufacturer s website for information, or

use the vendor and device IDs to cross-reference with /usr/include/hw/pci_devices.h.

You

can also search

CD-ROMs

You usually attach CD drives to a SCSI or EIDE ATA bus; which driver

you use depends on the bus.

Ensure that the hardware is set up correctly and that the BIOS

detects the hardware properly.

If you attached the CD drive to an EIDE bus, simply use the

devb-eide

driver.

If the drive is on a SCSI bus, you need to determine the proper driver for your

SCSI interface; see

Hard disks,

below.

The cam-cdrom.so shared

object provides a common access method for CD-ROM devices.

The drivers load the

cam-cdrom.so

and

fs-cd.so

shared objects by default.

CD-ROM devices support the ISO-9660 filesystem, and appear in

the /dev directory as /dev/cdx,

where x is the number of the drive, starting at 0.

Simply mount the drive using the

mount

utility, specifying cd as the type of filesystem:

mount -tcd /dev/cd1 /fs/cdrom

You don t need to remount the drive when you change CDs.

For information about specific options e.g. disabling audio extensions,

disabling multisession support, see

DVDs

Neutrino supports DVD ROMs and RAMs:

DVD ROM

You can treat DVD drives like CD-ROM drives. However,

Neutrino currently doesn t support UDF format DVD disks. You can use

DVD drives to read CDs, and the setup for a DVD ROM is the same

as for a CD-ROM. For more details, see

CD-ROMs,

above.

DVD RAM

You can treat DVD RAM drives like hard disks.

They appear in the /dev directory

as a CD, but you can mount and treat them just like a hard disk -- see

below.

Floppy disks

The driver for a floppy drive is

devb-fdc.

In order to use a floppy disk, you need to ensure that the

floppy controller is enabled in the BIOS, and that the BIOS is configured

to recognize the correct type of floppy drive e.g. 1.44MB/2.88MB.

The driver uses these locations as default:

I/O port 0x3f0

IRQ 6

DMA 2

If your controller is located at a different address, you can

change these locations in the driver s options.

The default cache size specified by

io-blk.so

is 15 of system memory, which is excessive for devb-fdc.

You ll probably want to reduce it to something more reasonable:

devb-fdc blk cache 128K

The driver creates a

/dev/fdx entry,

where x is the number of the floppy drive, starting at 0.

If no entry appears, the BIOS settings might be incorrect, or there

could be a problem with the controller.

Check the output from

sloginfo

for clues.

Once you have an entry in the /dev directory,

you need to mount the floppy disk.

The

command

detects the type of filesystem you re using e.g. DOS, QNX 4,

but you can also specify it on the command line.

To mount a DOS-formatted floppy disk, type:

mount -tdos /dev/fd0 /fs/dos_floppy

Use

mkdosfs

to format DOS floppy disks and DOS hard drives.

This utility supports FAT 12/16/32.

To mount a QNX 4-formatted floppy disk, type:

mount -tqnx4 /dev/fd0 /fs/qnx_floppy

You don t need to remount the drive when you change floppy disks.

Don t remove a floppy while the driver is still reading or writing data;

floppies are quite a bit slower than hard disks, so it can take a while.

Make sure the drive light is off.

Hard disks

A self-hosted system, by default, detects the disk controller that s installed

on the system, and then starts the appropriate driver for it.

On a self-hosted system, the

diskboot

utility in the OS image starts the block I/O drivers.

If you want to change the way that the driver is started, you ll

need to change the startup image and the options to diskboot.

For example:

diskboot -o devb-eide,blk cache 30m

Controlling How Neutrino Starts,

EIDE

EIDE interfaces use the

driver, which by default automatically detects the

interface and devices attached to it.

This driver includes support for UDMA Ultra Direct Memory Access modes, along

with the generic PIO Programmed Input/Output modes.

The supported hardware list includes adapters

and their supported features;

see the

introduction

to this chapter.

You can start the devb-eide driver without any options and,

by default, it

automatically detects the EIDE controller on the system:

devb-eide

When the driver starts, it detects all EIDE devices attached to the chain.

For each device, the driver creates an entry in the /dev

directory e.g. a hard drive appears as hdx,

where x is the number of the drive, starting from 0.

For example, suppose a system has two hard drives installed.

The driver creates the following

entries in the /dev directory:

/dev/hd0

Usually the primary master.

/dev/hd1

Usually the primary slave, or the next drive on the system the secondary

master.

If the system has one hard drive and a CD-ROM, the entries are:

The primary master.

/dev/cd0

The CD-ROM drive.

A slave drive must have a master drive.

When the driver starts, it displays on the

console the type of detected hardware, along

with other debugging information that gets sent to the system logger,

slogger.

To view the system log, run

sloginfo.

When you view the output from sloginfo, there will likely be a

number of ASC_MEDIA_NOT_PRESENT entries.

The driver logs these messages if there isn t a CD in the CD-ROM drive.

You can generally ignore them.

Troubleshooting for devb-eide

If the driver doesn t detect the interface or drives attached to it:

Check the supported-hardware part of our website to see if the interface

is supported; see the

introduction

to this chapter.

Even if your interface isn t listed as being supported, the EIDE controller

can work

in a generic mode that uses programmed input/output PIO modes, which is

slower, but works in almost all cases.

Ensure that the interface is correctly set up in the BIOS, and that the

BIOS can see the drives correctly.

Check that the drives are set up correctly; each slave drive must have a corresponding

master as per the ATAPI specs. A single chain can t have two master drives or two slave drives.

Ensure that the power connection is functioning correctly.

Pass the device ID and vendor ID to the driver.

Pass the I/O port and IRQ to devb-eide.

Here are some other problems that you might encounter and what you should

try:

If the driver hangs, disable busmastering e.g.

devb-eide eide nobmstr.

If you see sloginfo entries of:

eide_transfer_downgrade: UDMA CRC error downgrading to MDMA,

reduce the transfer mode and check the cables.

eide_timer: timeout path XX, device XX,

verify that the driver is using the correct interrupt, reduce the

transfer mode, and check the cables.

If a PCMCIA disk doesn t work when configured in contiguous I/O

mapped addressing, i.e. 0x320 not 0x1f0,

0x170,

specify the interface control block address.

The control block address is offset 12 from the base.

If a PCMCIA interface is located at I/O port 0x320 and IRQ 7,

specify:

devb-eide eide ioport 0x320:0x32c,irq 7,noslave

If your devices support UDMA 4 or higher, but sloginfo reports

that the driver is using a lower mode, make sure you re using

an 80-conductor cable.

If you have an 80-conductor cable and your devices support UDMA 4 or

higher,

but sloginfo reports that the driver is using a lower mode,

the device firmware might be out-of-date.

The driver relies on the device firmware to detect the cable type.

You can check to see if the device manufacturer has

a firmware upgrade or you can use the udma xxx

command-line option to override the mode.

For example:

devb-eide eide vid 0x8086,did 0x2411,pci 0,chnl 1,master udma 4

If the drives are detected, but they re running slowly:

sloginfo

to examine the devb- driver output in the system log.

It will tell you the current speed of the driver e.g.

max udma 5, cur udma 3.

Neutrino automatically uses the maximum UDMA mode, unless you ve specified

a maximum in the BIOS.

The following table shows the maximum mode and rate for each disk

specification.

The PIO, MDMA, and lower UDMA modes use a 40-pin cable;

higher UDMA modes require an 80-pin cable:

Specification

PIO

MDMA

UDMA 40-pin

UDMA 80-pin

Maximum rate

ATA

N/A

4 M/s

ATA 2

16 M/s

ATA 3

ATA 4

33 M/s

ATA 5

66 M/s

ATA 6

100 M/s

ATA 7

133 M/s

The maximum rate is the maximum theoretical burst interface throughput.

Sustained throughput depends on many factors, such as the drive cache size,

drive rotation speed, PCI bus, and filesystem.

Don t expect a UDMA-6 drive to have a sustained throughput of 100M/s.

Check to make sure that the device you re attempting to connect

can operate at the expected UDMA modes.

Correct the assignment of primary/secondary and master/slave interfaces.

For example, putting two hard drives as primary/secondary rather

than master/slave on the primary may allow driver parallelism.

SCSI devices

A SCSI Small Computer Systems Interface bus

is simply another bus that you can

attach multiple peripherals to.

Neutrino supports many brands and varieties of SCSI

adapters; see

Block-oriented drivers devb-

in the Summary chapter of the Utilities Reference.

When the SCSI driver starts up, it scans the bus for attached devices.

When the driver finds a supported device, it creates an entry in

the /dev directory e.g. a hard drive is

hdx, where x is the

number of the drive, starting from 0.

If the driver doesn t find any devices, it might not

know the device ID of the adapter.

Passing the device ID and vendor ID to the driver often corrects

this

problem.

On a self-hosted system, you can pass these options to the driver via

diskboot;

see

Controlling How Neutrino Starts.

In the following example, the driver automatically scans for SCSI devices on the chain and

adds them into the /dev directory as they re found.

For example, if the system has four hard drives in it, the entries in the

/dev directory are as follows:

/dev/hd0 -- lowest SCSI ID first

/dev/hd2

/dev/hd3 -- the last SCSI hard drive detected

When the driver starts, it sends debugging information to the system log,

which you can view using

This information is often very helpful when you re trying to

debug a problem with a SCSI adapter or device.

If the driver doesn t correctly detect a device, check the following:

Is the SCSI chain terminated correctly.

This is frequently the problem when a device doesn t

show up correctly, shows up and then disappears, or doesn t show up at all.

Is the SCSI adapter supported.

Even if an adapter claims to

be compatible with a supported adapter, that doesn t mean that the driver will

work with it correctly. Compatible doesn t mean identical.

To be certain, look for the device ID on our website; see

the

Does the SCSI BIOS see all the devices correctly.

If it does, then all the devices are set up correctly, and don t have any

conflicting SCSI IDs.

You can also check this by using another operating system;

if it detects the devices correctly and doesn t display any problems,

the setup is correct.

Remember that if a SCSI chain isn t terminated correctly, a device

may appear on the chain, but will likely have problems after some use. Each device on a

SCSI chain needs to have a unique ID number between 1 and the maximum value supported by the

adapter check the user manual for the adapter. If two devices have the same ID, one or

both may malfunction or not be recognized by the computer.

Is there a PCI-bridging problem. Try moving the SCSI card to another PCI slot.

Sometimes a PCI-bridging problem can prevent Neutrino from properly attaching to the card.

This can happen because Neutrino doesn t support bridges of type

other.

Is the BIOS set up for a PnP-aware OS. Neutrino isn t a PnP-aware OS.

Does the adapter or chain need an external power source.

If so, even if the device has power, it can t communicate

with your computer if the SCSI adapter doesn t have power.

Check the type of SCSI cable.

There are several types, and

the type of adapter you re using determines the type of cable you need.

Also check to make sure that there

are no bent pins on the cable. If you re using an adapter to convert between

SCSI 2 and SCSI 3, for example, make sure you re using an adapter that s

recommended for

your hardware. Not all adapters convert the connections correctly.

Under QNX 4, the SCSI drivers didn t support any device that had an ID

greater than 6.

This isn t a problem under Neutrino.

The maximum rate given for a SCSI device is the maximum theoretical burst

interface throughput.

Sustained throughput depends on many factors.

SCSI RAID

Currently, Neutrino supports only hardware RAID

Redundant Arrays of Independent Disks

devices.

There are many third-party solutions for SCSI RAID available for Neutrino;

search for them on the Internet.

LS-120

LS-120 is a SuperDisk drive that uses

new technology to greatly improve head alignment, enabling a

much greater storage capacity 120 MB than conventional 3.5-inch

disks.

Neutrino treats an LS-120 drive like an EIDE drive.

ORB

An ORB drive is a fast, large-capacity,

removable storage disk drive that uses 3.5 storage media and

attaches to the EIDE ATA chain. Ensure that the hardware is set

up correctly and that the BIOS detects the hardware properly. An

ORB drive is simple to set up, and appears in the /dev directory

as a hard disk. For example:

The hard disk as a primary master appears as /dev/hd0.

The ORB drive set up as a primary slave appears as /dev/hd1.

To mount an ORB drive:

mount /dev/hd1 /fs/orb_drive

You don t need to remount the drive when you change disks.

Zip and Jaz disks

Zip and Jaz

disks are large-capacity removable storage disks, used for backing

up hard disks and for transporting large files.

These disks attach to the EIDE ATA chain.

Before you attempt to use them, ensure that the hardware

is set up correctly and that the BIOS detects the hardware properly.

These drives are simple to set up, and they appear in the /dev

See also

Input drivers devi-

Determine which options are

appropriate for your setup, and then start the driver.

For example, here s how to start the driver for a Dynapro SC4 touchscreen:

devi-dyna dyna -4 fd -d/dev/ser1

This command starts the driver,

devi-dyna,

using the SC4 protocol -4, and a file descriptor that s

attached to serial port 1 fd -d/dev/ser1.

When you start the driver for the first time,

it returns an error stating that it can t read the calibration file.

To calibrate the touch screen, use the

calib

utility, while running Photon.

Audio cards

By default, the operating system detects your audio card. The enumerators

identify the card and use

io-audio

to start it. Audio

drivers in Neutrino are very simple to initialize. When you use io-audio,

you can use the -d option to pass the driver:

io-audio -vv -d audiopc

To see what other options you can use, see the documentation for

in the Utilities Reference and for your specific card.

If the operating system doesn t detect your card properly, you

can manually start the driver. In order to do this, you need to

identify the card.

You can find a list of supported hardware on our website; see

the

ISA cards

ISA cards are either Plug-and-Play or not.

You typically have to manually set up non-PnP ISA devices.

In order to identify your device, you need to have the manual for your

device or have a way to contact your device s manufacturer e.g. via

their website.

There isn t currently a Neutrino utility that lists the ISA devices that

are installed on a system.

Non-PnP-based

With non-PnP cards, you can manually start the driver and specify

the I/O port, IRQ, and DMA channel.

For example, this command starts the Sound Blaster driver:

io-audio -dsb ioport port,irq req,dma ch,dma1 ch

To find out what to set the I/O port and IRQ to, manually

open the system and look at the card. Then, start the driver using

the configuration settings that the card is set to.

Ensure that the I/O port and IRQ are reserved in the BIOS for non-PCI devices.

you re using a Sound Blaster card, check the following:

If the driver rejects the card, make sure that the I/O port doesn t

conflict

with another piece of hardware. Try changing the I/O port to see if

that helps.

If you hear a bit of sound and then nothing, make sure that the IRQ isn t

conflicting with another device and is reserved in the BIOS. You

can also try changing the IRQ as well.

If the driver starts correctly, but there s no sound,

check the DMA settings on the card and try changing them, if possible.

PnP-based

The device enumerator should configure and start ISA PnP cards.

If it doesn t,

you might need to obtain a copy of isapnp, which is

used to initialize ISA PnP cards. Neutrino doesn t supply this utility,

but it s freely available on the Internet and has been ported

to Neutrino.

PCI Cards

The device enumerator should start PCI cards correctly.

If your PCI card doesn t work, swap PCI slots. Sometimes the

IRQ that s assigned to the particular slot doesn t work well with

the card.

For additional information about the card, use the

utility.

For a list of supported hardware, see our website, as described in

PCCARD and PCMCIA cards

Neutrino supports PCMCIA 1.0/2.0 and CardBUS type cards.

By default, the driver detects the

ISA/PCI based controller. If an adapter isn t detected, check the

supported hardware page to ensure that your PC Card

adapter s chipset is supported.

Currently the driver doesn t let you specify the adapter s I/O port

and IRQ, but you can specify the card s I/O port and IRQ.

If the driver fails to start:

Ensure that the

devp-pccard

server has a free memory window at 0xD4000.

Check the BIOS on the PC or Laptop to see that this memory

isn t cached or used by another device.

Check that the PC Card controller in the BIOS is set to CardBus/16bit,

not PCIC mode.

If the chipset is set up in PCIC compatible mode, the chip works

like an Intel 82365-compatible PCMCIA controller and

isn t visible in the PCI space. If the chipset

is set to CardBus/16bit, the chip is visible in the PCI

space and operates as a PC Card adapter.

To display PC Card information, use the

The output that appears on your screen should look like this:

Sock Func Type Flags PID Base Size IRQ

1 Empty -----MF------ None

2 0 Network C---I- ------ None 0x300 32 7

2 Empty ----MF--------- None

Each socket has two entries because the driver devp-pccard

supports combination cards that give room for two functions in each slot.

The categories displayed in the output example above are:

Sock

The slot where the PC Card is attached. In the example

above, the Network card appears in slot 2.

Func

Used when the card is a multifunction PC Card.

Type

A label for the PC Card s function. If the card is a

Network card, the Type column displays Network.

Flags

Flags that aren t set are marked as -.

The following table lists possible set flags:

This flag:

Has a set value of:

Card in

Battery low

Scheduled to be configured

Not enough resources to configure card

I or M

I/O card or memory card

Not configured

Window is part of previous configuration

Window is an unlockable window

Window is a temporary window

Machine booted from this device

X or W

Locked exclusive / locked read/write

Locked read-only

Level-mode IRQs

Shared IRQs

Attribute memory

Wide 16-bit memory access

PID

The process ID of the process attached to the PC Card driver

devp-pccard.

Base

The base address of the PC Card. This information is

useful for starting device drivers.

Size

The number of bytes in the I/O port range.

IRQ

The PC Card s IRQ. This information is useful when starting

the driver manually.

USB devices

A Universal Serial Bus USB provides a hot-swappable, common interface for

e.g network, input, character I/O, audio, and hubs.

For more information on USB, USB specifications, and a list of frequently

asked questions, see

www.usb.org.

If you don t know what kind of USB device you re using, you can use the

usb

usb -vvv less

The output from this command looks like this:

Device Address : 1

Vendor : 0x05c7 QTRONIX

Product : 0x2011 USB Keyboard and Mouse

Device Release : r1.12

USB Spec Release : v1.00

Serial Number : N/A

Class : 0x00 Independent per interface

Max PacketSize0 : 8

Languages : 0x0409 English

Current Frame : 511 1024 bytes

Configurations : 1

Configuration : 1

Attributes : 0xa0 Bus-powered, Remote-wakeup

Max Power : 50 mA

Interfaces : 2

Interface : 0 / 0

Class : 0x03 HID

Subclass : 0x01 Boot interface

Protocol : 0x01 Keyboard

Endpoints : Control 1

Endpoint : 0

Attributes : Control

Max Packet Size: 8

Endpoint : 1

Attributes : Interrupt/IN

Interval : 20 ms

Interface : 1 / 0

Protocol : 0x02 Mouse

Max Packet Size: 8

The vendor and product fields indicate the type of device, and possibly what

chipset it uses.

The common types of USB controllers are:

UHCI

Universal Host Controller Interface.

EHCI

Enhanced Host Controller Interface.

OHCI

Open Host Controller Interface made by others.

The EHCI controller supports high speed signalling only.

Either a OHCI or UHCI controller s should be present to

support low- or full-speed devices.

If your system doesn t have an EHCI controller, the

device will work at the slower speed.

The operating system needs to run the stack in order to know

how to interact with USB devices and controllers.

You need to:

Identify your controller.

The documentation for the hardware should describe the type

of controller OHCI, UHCI, or EHCI.

If you don t know what type of controller you re

using, you can identify it using:

pci -vvv

Find the entry for the USB controller to determine

the manufacturer/vendor ID and device ID. You can either find

the information on the manufacturer s website

www.usb.org,

or use the vendor and device IDs to cross-reference it at

The class codes that appear in the output from pci -vvv are:

Class Code

Controller Type

0c0300

0c0310

0c0320

There might be multiple chips and therefore multiple drivers that you need

to load.

You can also try running just one of the USB stacks;

if it fails, try running another stack.

io-usb

stack with the appropriate module:

This should create an entry in /dev called

/dev/io-usb/io-usb.

If you re starting the USB stack

and a driver in your startup scripts,

make sure that you use the

waitfor

command to make sure that /dev/io-usb/io-usb has appeared

before you start the driver.

io-usb -dohci

waitfor /dev/io-usb/io-usb

devu-prn

When the stack is running, start the device drivers, as described

USB hubs don t need a driver; the stack itself supports them.

Printers

For a USB printer, start the USB stack, and then

devu-prn.

Photon supports USB Human-Interface Devices HID such as keyboards and mice.

To connect USB HIDs:

Start the USB stack, as described above.

Start

io-hid,

loading the

devh-usb.so

module:

io-hid -dusb

If your system also uses serial or PS/2 input devices, you can load the

devh-ps2ser.so

module as well.

After starting Photon, start

devi-hid

for the USB HID devices as follows:

devi-hid kbd -u USB_device_number mouse

You can start io-hid in your rc.local file,

but not devi-hid,

because Photon hasn t started when your system runs rc.local.

Instead, add the devi-hid command to the input trap file; see

In Photon, once the devices are running, you can use the

utility

to configure the mouse.

From the shelf, click

You can use the

hidview

utility to get information about the human-interface devices.

For USB touchscreens, start the USB stack, then

io-hid,

loading the

devh-usb.so

Then, start

devi-microtouch:

devi-microtouch microtouch touchusb

Ethernet adapters

For Ethernet adapters, start the USB stack, then

io-net,

loading the appropriate driver.

For example, to start the driver for a Kawasaki-based USB Ethernet adapter,

do the following:

io-net -dklsi options

Mass-storage devices

devb-umass

driver supports devices that follow the

Mass Storage Class Specification.

You can determine that the device is suitable by looking for the following

information in the output from usb -vv:

Mass Storage Class 08h

SubClass Code Command Block Specification

01h Reduced Block Command RBC

02h SFF-8020i, MMC-2 ATAPI

04h UFI

05h SFF-8070i

06h SCSI transparent

Protocol Code Protocol Implementation

00h Control/Bulk/Interrupt

with command completion interrupt

01h Control/Bulk/Interrupt

with no command completion interrupt

50h Bulk-Only Transport

To use a USB mass-storage device on a Neutrino system, start

io-usb

as described above, then the devb-umass driver.

By default, this driver creates an entry for disk-based devices

in /dev in the form

/dev/hdn, where n is the drive number.

Once you ve started the driver, you can treat the device like a disk.

For example, for a mass-storage device that uses the UHCI controller, type:

io-usb -d uhci

devb-umass cam pnp

Troubleshooting

No device is created in /dev.

The device might not conform to the Mass Storage Class Specification.

Check the output from usb -vv.

No fdn device was created in

/dev for a floppy drive.

The default name is /dev/hdn.

You can use the name command-line option to

cam-disk.so

to override the prefix.

Character devices

General serial adapters

By default, a serial port driver automatically detects the I/O port and IRQ.

A standard PC system uses the

devc-ser8250

driver; the BSP documentation

indicates the drivers specific to your target hardware.

If the driver doesn t detect all the serial ports, ensure that the ports are

enabled in the BIOS.

If the ports are enabled, try specifying the I/O port and

IRQ of the ports when you start the driver.

Use a comma to separate the I/O port and the IRQ; use a space to separate

each port-IRQ pair in the command.

devc-ser8250 3f8,4 2f8,3

If you start a serial driver for a UART or modem when another serial driver

is already running, you need to use the -u option to give the

new driver a number to append to the device name so that it doesn t

conflict with any existing /dev/ser entry.

The standard devc-ser8250 driver supports only the

RS-232 protocol.

The Character Driver Development Kit DDK includes

the source to devc-ser8250,

which you can use to implement any additional protocols or features.

The serial drivers support software and hardware flow control:

To enable software flow control, start the serial driver with the

-s option, or use

stty

after starting the driver:

stty osflow isflow /dev/ser1

To disable software flow control, start the driver with the

-S option, or use:

stty -osflow -isflow /dev/ser1

To enable hardware flow control, start the driver with the

-f option, or use:

stty ohflow ihflow /dev/ser1

To disable hardware flow control, start the driver with the

-F option, or use:

stty -ohflow -ihflow /dev/ser1

In edited mode -e, flow control is disabled.

Don t enable software and hardware flow control at the same time.

Heavy serial port usage can be very taxing on some systems;

by default, the serial adapter triggers an interrupt for each character

transmitted or received.

You can use these options to reduce the number of interrupts:

-T number

Enable the transmit FIFO and set the number of characters to be

transmitted at each TX interrupt to 1, 4, 8, or 14.

The default is 0 FIFO disabled.

-t number

Enable the receive FIFO and set its threshold to 1, 4, 8, or 14 characters.

The default is 0 trigger disabled.

A receive timeout guarantees that the characters won t remain buffered too

long.

For example, imagine that the device receives:

This sentence is coming across the serial port.

By default, the system has to service 47 interrupts to receive this sentence.

If you set the receive trigger level to 14,

the number of interrupts is reduced to four.

This helps the overall system performance, but you re trading off reliability;

the higher the receive trigger -t, the higher the possibility

of losing data.

Multiport serial adapters

For multiple serial adapters, you may need to specify the I/O port

and IRQs manually in the driver for each port see

General serial adapters

for examples. By default, the driver should detect the ports and IRQs,

but with some multiport adapters,

the enumerators don t detect the ports correctly.

You can edit the

enumerators to detect your multiport card and have it set up each

port for you. You need to edit the

/etc/system/enum/devices/overrides file;

chapter in this guide,

Parallel ports

On a standard PC and some x86 systems, parallel ports use the

devc-par

driver; see the BSP documentation for the driver for your target hardware.

By default, the driver detects the parallel port. If you need

to, you can use the -p option to specify the location

of the parallel port.

If the driver fails to detect your parallel port, ensure that

the port is enabled in the BIOS. If that fails, try specifying

the I/O port when you start the driver.

Terminals

On a standard PC and some x86 systems, the

devc-con

driver controls the physical console, which consists of the

display adapter, the screen, and the system keyboard.

By default, the driver is configured for up to four virtual

consoles, /dev/con1 /dev/con4.

The devc-con driver is also the keyboard driver for

non-USB keyboards in text mode. You can start the driver with this

command:

devc-con

I/O attributes

To set or display the I/O attributes for a character device tty, use the

stty

For more information about setting up your terminal, see

Terminal support

in Using the Command Line.

Network adapters

The main steps in setting up a network adapter are:

identifying your Network Interface Card NIC

starting the driver

making sure the driver and hardware communicate

Identify your NIC

The documentation for the hardware should describe the type of

chipset used.

If you don t know what type of chipset you re using, you can

identify it using pci -vvv.

Find the entry for the Network controller and it ll give you

details on the manufacturer/vendor ID and device ID. Either find

the information on the manufacturer s website, or use the vendor

ID and device ID to cross-reference it with this online site:

With the information you get from that site, you can visit the

QNX supported hardware site; see

In the Network section, locate your chipset and its

associated driver.

Start the driver

Once you ve located the correct driver for your hardware, use

io-net

to start the driver. You can either

start the driver as an option to io-net, or you can

mount the driver into an already running copy of io-net.

For example, to start io-net with the

devn-el900.so

3Com 905 module, type:

io-net -d el900 -t tcpip

To mount the module, type:

io-net -t tcpip

mount -T io-net devn-el900.so

The driver automatically detects similar network adapters

for multiple networks. You can use the mount utility to mount different

adapters.

Make sure the driver is communicating properly with the hardware

Use the

nicinfo

utility to check if you re receiving

and sending packets. If you aren t receiving packets on a high-traffic

network, the driver and the hardware might not be communicating.

Here s some typical output from this command:

Physical Node ID 000102 C510D4

Current Physical Node ID. 000102 C510D4

Current Operation Rate 100.00 Mb/s full-duplex

Active Interface Type. MII

Active PHY Address. 3

Power Management State Active

Maximum Transmittable data Unit 1514

Maximum Receivable data Unit 1514

Receive Checksumming Enabled TCPv6

Transmit Checksumming Enabled. . TCPv6

Hardware Interrupt. 0x5

DMA Channel. . 0

I/O Aperture. 0xd400 - 0xd47f

ROM Aperture. 0

Memory Aperture. 0xe6000000 - 0xe6000FFF

Promiscuous Mode Off

Multicast Support. . Enabled

Packets Transmitted OK 104

Bytes Transmitted OK. . 10067

Broadcast Packets Transmitted OK. . 6

Multicast Packets Transmitted OK. . 1

Memory Allocation Failures on Transmit. . 0

Packets Received OK 1443

Bytes Received OK. . 168393

Broadcast Packets Received OK. . 427970

Multicast Packets Received OK. . 37596

Memory Allocation Failures on Receive 0

Single Collisions on Transmit. . 0

Multiple Collisions on Transmit 0

Deferred Transmits. 0

Late Collision on Transmit errors. 0

Transmits aborted excessive collisions 0

Transmits aborted excessive deferrals . 0

Transmit Underruns. 0

No Carrier on Transmit 0

Jabber detected. 0

Receive Alignment errors. 0

Received packets with CRC errors. . 0

Packets Dropped on receive. . 0

Ethernet Headers out of range. . 0

Oversized Packets received. . 0

Frames with Dribble Bits. 0

Total Frames experiencing Collision s . . 0

The output from nicinfo depends on what the driver supports;

not all fields are included for all drivers.

However, the output always includes information about the bytes and

packets that were transmitted and received.

The categories shown in the above example are described below.

When dealing with a network problem, start with these:

Physical Node ID

The physical node ID is also known as the Media Access Control

MAC address. This value is unique to every network card, although

some models do let you assign your own address. However, this

is rare and generally found only on embedded systems.

If the value represented is

FFFFFF FFFFFF

000000 000000,

there s likely something wrong with the setup of the hardware,

or you need to assign a MAC address to the card. Check

the hardware manual to see whether or not this is the case.

If the hardware didn t get set up correctly, the MAC address

may not always appear as shown above.

The first six digits of the MAC address are the vendor ID.

Check the entries against the list at

see if the vendor ID is valid. Then check the card ID the last 6 digits.

The card ID should be something semi-random. A display similar to

444444 is likely incorrect.

Current Physical Node ID

The current physical node ID is shown if a card has been set

up to spoof the ID of another card. Basically, a

parameter is passed to the driver telling it that the node s ID

is actually the value that appears. Depending on the card, some

drivers will accept this. What spoofing does on a higher software

level, is filter out the packets that were meant for this node ID.

This method is considerably slower than if you let the card filter

out the packets on a hardware level. Because the card is set in promiscuous

mode, it has to accept all packets that come in and use a software

mode to sort them.

Another way of thinking about this is to compare

it to a postal system, where if we wanted to pretend to

be someone else, we would accept all mail from the Post Office.

However, we would then have to sort all the mail. This would take

a much longer time compared with the amount of time the Post Office

would take to presort the mail, and give us only the mail addressed

to us.

Promiscuous Mode,

Current Operation Rate

The media rate is the speed at which the network card operates.

On most cards, it s either 10Mb/s or 100 Mb/s. This display

also shows what form of duplex the card uses. Most cards run at

half or full-duplex transmission:

Full-duplex transmission means that data can be transmitted

in both directions simultaneously.

Half-duplex data transmission means that data can be transmitted in

both directions, but not at the same time.

The easiest way to illustrate this is to think of a road.

If the road has two lanes, it s full-duplex, because cars can drive in both

directions at the same time without obstructing the other lane. If the road

has only a single lane, it s half-duplex, because there can be only one car on

the road at a time.

When you examine the media rate, check the speed, the form of duplex, and

what the hub supports.

Not all hubs support full-duplex.

Active Interface Type

This is the type of interface used on the Ethernet adapter.

This is usually UTP unshielded twisted pair, STP shielded

twisted pair, Fiber, AUI Attachment Unit Interface, MII, or BNC

coaxial.

Active PHY Address

This is an identifier that tells you which of the physical

PHYs were used to interface to the network. The numbers

range from 0 - 31 and change, depending on

whether or not you specified a specific PHY or if you let the

driver select the default which varies from card to card.

Power Management State

This value tells you the NIC s current power status:

Off, Standby, Idle, or Active.

If you can t send or receive packets, make sure the status

is Active; if it isn t, there may

be a problem with power management on your system.

Maximum Transmittable data Unit

The Maximum Transmittable data Unit MTU is the size of the largest

frame length that can be sent on a physical media. This isn t

commonly used for debugging; however, it may be useful for optimizing

a network application. A value of 0 is invalid and is a

good indicator that the card isn t set up correctly.

The default value is 1514.

Maximum Receivable data Unit

This is the MTU s complement; it affects the largest frame length

that can be received. The default value is 1514.

Receive Checksumming Enabled, Transmit Checksumming Enabled

Not all cards support these options.

If your adapter supports them, they tell your card which

check-summing method to use: IPv4, TCPv4, UDPv4, TCPv6, or UDPv6.

Hardware Interrupt

The hardware interrupt is the network card s interrupt request

line IRQ. How an IRQ is assigned depends on whether the card

is a PCI or an ISA card. In the case of a PCI card,

pci-bios

assigns

the IRQ; for an ISA card, the IRQ is hard-wired.

Two ISA devices can t share the same IRQ, but two PCI devices can.

DMA Channel

This is the DMA channel used for the card.

This varies, depending on the card and on the channels it has available.

I/O Aperture

The I/O aperture is a hexadecimal value that shows the address in I/O space

where the card resides.

The I/O aperture uses the I/O address between the given values

to locate and map the I/O ports.

The range depends on the platform.

Memory Aperture

The memory aperture is a hexadecimal value that shows the address in

memory where the card s memory is located.

The memory aperture uses the memory address between the given values to

locate and map memory.

ROM Aperture

The ROM aperture is a hexadecimal range that shows the address of the

card s ROM.

The ROM aperture uses the memory address between the displayed values

to locate and map memory.

Promiscuous Mode

When a card is placed in promiscuous mode,

the card accepts every Ethernet packet sent on

the network. This is quite taxing on the system but is a common

practice for debugging purposes.

Also, when a card is placed in

promiscuous mode, a network MAC address can be spoofed,

i.e. the card accepts all packets whether they re addressed

to it or not. Then on a higher software level, you can accept packets

addressed to whomever you please. Promiscuous mode is disabled by

default.

Multicast Support

When you enable multicast mode, you can mark a packet with a

special destination, so that multiple nodes on the network may receive

it. Multicast packets are also accepted.

Packets Transmitted OK

Before you look at this value, determine that some form of network transfer

ping,

telnet,

file transfer was attempted.

If a card isn t set up properly, the number of sent packets shown

here is either very small or zero. If the card

isn t displaying any sent packets, the cause is probably a driver

problem. Check all the options you re passing to the driver; one

or more may be incorrect.

Bytes Transmitted OK

This is the number of bytes of data sent on the network.

This value increases with the number of packets transmitted on the

network.

Total Packets Transmitted Bad

You can use this statistic to determine if you have faulty hardware.

If all the sent packets are reported as bad,

there s likely a hardware problem, but you might be using the wrong driver.

Check the

hardware for compatibility. If it looks as if it s hardware-related,

try switching the hardware to see if the problem disappears.

Broadcast Packets Transmitted OK

This is the number of broadcast packets transmitted from the NIC.

Multicast Packets Transmitted OK

This is the number of multicast packets transmitted from the NIC.

Memory Allocation Failures on Transmit

Before transmitting data, the driver reserves system memory for

a buffer to hold the data to be transmitted. Once the

card is ready, the buffer is sent to it.

When a memory-allocation error occurs, the system is likely very low on memory.

Make sure that there s sufficient memory on the system; if

you continuously get this error, consider adding

more memory. Another thing to check for is memory leaks on the

system, which may be slowly consuming system memory.

Packets Received OK

This value states how many packets were successfully received

from the network card. If a card is having problems receiving data,

check the cables and the hub connection. Problems receiving data

might be related to the driver.

It s possible the driver can be properly

set up and able to send data, but may not be able to receive. Usually

when data is received but doesn t get sent, the driver is the cause.

Check the driver s setup to make sure it s initialized correctly.

to check the system log for clues.

Bytes Received OK

This is the number of bytes of data received from the network.

This value increases with the number of packets received.

Single Collisions on Transmit

This is the number of collisions that were encountered while

trying to transmit frames.

The NIC checks

for a carrier sense when it knows that the network hasn t been used

for a while, and then starts to transmit a frame of data.

The problem occurs when two network cards check for the

carrier sense and start to transmit data at the same time.

This error is more common on busy networks.

When the NICs detect a collision, they stop transmitting and

wait for a random period of time.

The time periods are different for each NIC, so in theory, when the wait time

has expired, the other NIC will have already transmitted or will

be still waiting for its time to expire, thus avoiding

further collisions.

You can reduce this type of problem by introducing a full-duplex network.

Multiple Collisions on Transmit

This error is due to a attempted transmission that has had

several collisions, despite backing off several times.

This occurs more frequently on busy half-duplex networks.

If there are a lot of these errors, try switching to a full-duplex network,

or if the network is TCP/IP based, try introducing a few switches

instead of hubs.

Deferred Transmits

Commonly found on half-duplex networks, this value doesn t mean that there

are problems.

It means that the card tried to

send data on the network cable, but the network

was busy with other data on the cable. So, it simply

waited for a random amount of time. This number can get high if

the network is very busy.

Late Collision on Transmit errors

Late-collision errors that occur when a card has

transmitted enough of a frame that the rest of the network should

be aware that the network is currently in use, yet another system

on the network still started to transfer a frame onto the line.

They re the same as regular collision errors, but were

just detected too late.

Depending on the protocol, these types of errors can be

detrimental to the protocol s overall throughput. For example, a

1 packet loss on the NFS protocol using the default

retransmission timers is enough to slow the speed down by approximately 90.

If you experience low throughput with your networking,

check to make sure that you aren t getting these types of errors.

Typically, Ethernet adapters don t retransmit frames that have been

lost to a late collision.

These errors are a sign that the time to propagate the signal

across the network is longer than the time it takes for a network

card to place an entire packet on the network.

Thus, the offending

system doesn t know that the network is currently in use, and it

proceeds to place a new frame on the network.

The nodes that are

trying to use the network at the same time detect the error

after the first slot time of 64 bytes. This means that the NIC detects

late collisions only when transmitting frames that are longer than 64 bytes.

The problem with this is that, with

frames smaller than 64 bytes, the NIC can t detect the error.

Generally, if you experience late collisions with large frames on

your network, you re very likely also experiencing late

collisions with small frames.

These types of errors are generally caused by Ethernet cables that are

longer than that allowed by the IEEE 802.3 specification, or are

the maximum size permitted by the particular type of cable, or

by an excessive amount of repeaters on the network between

the two nodes.

Another thing to note is that these errors may actually be caused

by a node on the network that has faulty hardware and is sending

damaged frames that look like collision fragments.

These damaged frames can sometimes appear to a network card to be

a late collision.

Transmits aborted excessive collisions

This error occurs if there are excessive collisions on the

network. The network card gives up on transmitting the frame

after 16 collisions. This generally means that the network is

jammed and is too busy.

Routers also give up on transmitting a frame if they experience

excessive collisions, but instead of alerting the original

transmitter, routers simply discard the frame.

If these sort of errors are being experienced, see if the

network can be reduced, or introduce a strategically placed switch

into the network to help eliminate the number of packets that are

being placed on the entire network.

Switching to a full-duplex network also resolves these problems.

Transmits aborted excessive deferrals

Aborted transmissions due to excessive deferrals mean that the NIC

gave up trying to send the frame, due to an extremely busy

You can resolve this type of problem by switching to a full-duplex network.

Transmit Underruns

Chips with a DMA engine may see this error. The DMA engine copies

packet data into a FIFO, from which the transmitter puts the

data on the wire. On lower-grade hardware, the DMA might not be

able to fill the FIFO as fast as the data is going on the wire, so

an underrun occurs, and the transmit is aborted.

No Carrier on Transmit

When the NIC is about to transfer a frame, it checks first to

make sure that it has carrier sense much like before you dial the

phone, you check to make sure you have a dial tone. While the NIC

is transmitting the frame, it listens for possible collisions

or any errors. These errors occur when a NIC is transmitting a

frame on the network, and it notices that it doesn t see its own

carrier wave much like when you are dialing a number on the phone

and you can hear the dial tones being pressed.

These errors are caused by plugging and unplugging cables on the

network and by poor optical power supplied to the

Fiber Optic Transceiver FOT.

Jabber detected

You typically see this error only on a 10Mbit network.

It means that a network card is continuing to transmit after a packet

has been sent. This error shouldn t occur on faster networks, because

they allow a larger frame size.

Receive Alignment errors

A receive-alignment error means that the card has received a

damaged frame from the network. When one of these errors occurs,

it also triggers an FCS Frame Check Sequence error.

These errors occur if the received frame size isn t a

multiple of eight bits one byte.

These errors are commonly due to faulty wiring, cable runs that

are out of the IEEE 802.3 specification, a faulty NIC, or

possibly a faulty hub or switch. To narrow down this problem, do a

binary division of the network to help eliminate the source.

Received packets with CRC errors

An entry in this field indicates the number of times, on a hardware

level, the card received corrupt data.

This corruption could be caused by a faulty hub, cable, or network card.

The best way to try to solve Cyclic Redundancy Check CRC

errors is to do a binary division of the

systems on the network to determine which system is sending bad

data. Once you ve done that, you can start replacing the hardware

piece by piece. Because this error is on the receiving end, it s

difficult to determine if the CRC is bad on a sent packet.

Packets Dropped on receive

This usually means you got an overrun while receiving a packet. This

has to do with DMA and the FIFO, like a Transmit Underrun, except

in this case, the DMA engine can t copy the packet into memory as

fast as the data is coming from the network, and the packet gets

dropped.

Like the Transmit Underrun, this is generally due to poor

hardware.

Ethernet Headers out of range

This entry indicates the number of packets whose Ethernet type/length

field isn t valid.

Oversized Packets received

An oversized packet is simply a received packet that was too big

to fit in the driver s Receive buffer.

Frames with Dribble Bits

Dribble bits are extra bits of data that were received after the

Ethernet CRC.

They re commonly caused by faulty hardware

or by Ethernet cabling that doesn t conform to the 802.3

specifications.

Total Frames experiencing Collision s

This is the total number of frames that have experienced a

collision while trying to transmit on the network. This can

sometimes be high, depending on how busy the network is.

A busy network experiences these types of errors more often than

a quiet one.

Modems

You can have any of the following types:

Internal ISA Plug-and-Play or not

PCI-based

External

Cable

Internal modems

Internal modems can be ISA

and are either Plug-and-Play PnP or not.

You have to manually set up non-PnP ISA devices.

In order to identify your device, you need to have the documentation

for the device, or be able to contact the device manufacturer to

have it identified. Currently, there is no utility within Neutrino to

obtain a list of ISA devices installed on your system.

ISA non-PnP

Configure the modem to use an I/O port and IRQ that don t conflict

with anything else in the system.

driver should autodetect the

modem and it should appear in the /dev directory as

serx, where x is an integer.

There may be more than one entry under the name. Assume that

the first two entries represent the comm ports of the system.

Any additional entry is likely the modem.

If in doubt, try all ser entries

with

qtalk.

Testing Modems,

Entries will usually appear in this fashion:

Comm1 is enabled in the BIOS

Comm2 is disabled

Modem is configured to Comm2 s ioport and IRQ

In the /dev directory you ll see:

ser1 -- Comm1

ser2 -- Modem

ISA PnP

If you have an ISA PnP modem that can be manually

assigned an IRQ and I/O port via jumpers, we recommend that you use the manual

method rather than Plug-and-Play.

driver should automatically detect the

modem, which should appear in the /dev directory

as serx, where x is an integer.

There may be more than one entry in /dev under

the name ser.

Assume that the first two represent

the comm ports of the system. Any additional entry is likely

the modem. However, if in doubt, try all ser entries

If the modem isn t detected, seek out the isapnp utility

to configure the modem s I/O port and IRQ, and then specify them when you

start devc-ser8250.

PCI-based modems

If no entry is created, check the output from pci -vvv and

see what I/O port and IRQ are assigned to the modem. Use the correct

I/O port and IRQ from pci -vvv to start devc-ser8250.

When you use the appropriate I/O port and IRQ, the /dev directory

entry gets created for you.

External modems

External modems are easy to set up. Look in the /dev directory

for the serial port that the modem is attached to. You ll attach

this at the back of the system. If you know the modem is attached

to serial port 1, then look in the /dev directory

for ser1.

Cable Modems / ISDN

We assume that your cable modem is attached to your system via

a network card and that the driver for your card has been started and

is running properly.

If this isn t the case, see

Network adapters,

earlier in this chapter.

To set your configuration, use the TCP/IP configuration tool.

phlip

on the command line, by selecting Network from Photon s shelf, or by

choosing

Launch-- Configure-- Network.

Then do the following:

Go to the Devices tab and use these settings:

Choose en0.

Choose DHCP for the connection.

In the Server field, enter the machine ID given by

the cable network operator.

Go to the Network tab and use these settings:

Use the machine ID as the hostname.

Set the domain name to be the domain name of your cable network operator.

Set the gateway to be the IP address of your cable network operator.

The default netmask 0.0.0.0 is filled in automatically.

In the Name Servers field, add any IP addresses

you know there may be more than one.

The first entry should be

the IP address of your cable operator.

Reboot your machine. DHCP will start automatically.

Testing modems

You can use

qtalk

to test your modem:

Make sure the modem is plugged into the phone line.

command to set the modem s baud rate.

For example, to set the speed of the modem on /dev/ser1

to 57600 56K modems use this speed, type:

stty baud 57600 /dev/ser1

Type qtalk -mdevice, where device

is the name of the serial port e.g. /dev/ser1.

Type at.

The modem should reply OK.

Troubleshooting modems

If you followed the instructions above,

but the modem doesn t reply OK, check the following:

Make sure your baud rate settings are correct.

Is the modem plugged in.

Is the modem a software modem.

Neutrino doesn t support Win modems or HSP Host Signal Processor modems

otherwise known as soft modems.

Neutrino works with PnP modems, but

you must specify in the BIOS that you aren t running a PnP-aware OS.

Does the modem conflict with another device at the same I/O port

and IRQ.

If the modem is an internal ISA modem, you may need to reserve an I/O port

range and IRQ in the BIOS so that the PCI doesn t use it.

Have you disabled the comm port in the BIOS if you re using the

same I/O port and IRQ of a comm port. This applies only to internal

modems.

Video cards

Our website includes a list of the video cards Neutrino supports; see

Automatically detecting your card

By default, the operating system uses the

crttrap

utility to detect your video card.

We provide crttrap as a convenience utility;

in a static hardware environment embedded board, you shouldn t need it.

Ensure you have a graphics-modes file or the correct

command line in your startup scripts to start your video card.

Don t run crttrap while you re running Photon.

The crttrap program queries the card to get all of its possible

configurations and creates a file called

/etc/system/config/graphics-modes

that you can edit.

The first line in this file is the one that io-graphics uses.

For information about the different graphics options, see

io-graphics

In an embedded system, you don t need to use crttrap to get

the configuration, because it isn t likely to change.

Instead, you can call io-graphics directly, specifying the

appropriate options.

io-graphics -dl devg-rage.so options.

Neutrino s default setup starts the graphics driver for you via

the ph script located in /usr/bin.

Changing video modes in Photon

To change the video modes of your graphics adapter in Photon,

use the Display Configuration tool.

phgrafx

on the command line, by selecting Graphics from the shelf, or by choosing

Launch-- Configure-- Graphics.

To change the video modes:

Select the configuration that you want to use.

Click Apply.

The screen turns black, and then should go into the new mode.

If the mode you selected didn t work properly, you can wait for 15 seconds,

at which point the display automatically reverts to the previous settings,

or you can press Esc or Enter.

When the mode has changed and seems to be working properly, click Accept.

This utility also lets you change the driver, resolution, refresh rate,

and palette 8-bit color mode only, and disable the hardware cursor.

If you want to edit the command line that s used to start the adapter,

click the Advanced button.

This is the same as editing the

top line in the graphics-modes file.

Manually setting up your video card

Identify your video adapter. The documentation for the hardware

should describe the chipset used on your adapter.

Identify which driver you should be using.

Use the list of supported hardware to determine which driver is

appropriate for your adapter; see

Test the driver by using

io-graphics

to start it manually:

io-graphics -drage

vid 0x1002,did 0x4755,index 0,xres 1024,photon,yres 768,bitpp 16,refresh 80

-pphoton

For information about the options, see

Graphics drivers devg-

in the Utilities Summary in the Utilities Reference, as

well as the entry for

io-graphics.

You ll also need to start Photon and other utilities to ensure

that this is working correctly. The best thing to do is create

a script that starts Photon. See the

script for

ideas and examples.

You can also manually add this command to the graphics-modes

file at

the top line to test it out.

Setting up multiple displays

configuration file to

set up a dual- or multiple-monitor display.

You can configure either

two or more separate cards for multiple displays, or a

single device to support multiple displays. These graphics device

drivers support multiple displays:

devg-radeon.so

Graphics driver for ATI RADEON chipsets

devg-matroxg.so

Graphics driver for Matrox Millenium G-series chipsets

For specific chipsets and the number of displays supported, see the

documentation for

and devg-matroxg.so

For information about the io-graphics configuration

file format and the options you can set, see

The io-graphics configuration file

in the documentation for io-graphics in the

Utilities Reference.

utility or io-graphics command line

to configure only single displays, not multiple displays.

We don t guarantee support for arbitrary combinations of video cards.

Here s a step-by-step example of creating a configuration file

for a Radeon 9700 Pro video card for dual-headed display:

Create the GLOBAL section. This has a single

entry, devices, that lists all the video cards supported.

In this example, there s one device:

GLOBAL devices radeon

For each device, create a DEVICE.name section.

This section configures the hardware-specific settings. In this example,

we set the:

graphics driver DLL, which is typically in the form devg-device_name.so,

or in this case devg-radeon.so

hardware vendor and device ID and the PCI index; see

PCI/AGP devices,

earlier in this chapter

number of displays supported by the device; 2 in this case

plugins, which in this case is the Photon plugin

Photon server, which we leave blank to use the default Photon

server specified by the PHOTON environment variable

/dev/photon by default

DEVICE.radeon

dllpath devg-radeon.so

pci_vendor_id 0x1002

pci_device_id 0x4e44

pci_index 0

displays 2

plugins photon

photon

For each display, create a DEVICE.name.number section.

These sections configure each display. In this example, each display

has the same resolution 1024 768, color depth 16 bit color,

and refresh rate 60 Hz. Notice that the second display Photon

region is offset by 1024, the width of the first display.

DEVICE.radeon.0

xres 1024

yres 768

bitpp 16

refresh 60

DEVICE.radeon.1

region_x 1024

For each plugin, create a PLUGIN.name section.

In this example, we set the Photon plugin DLL:

PLUGIN.photon dllpath gri-photon.so

Save the configuration file, in this case as /usr/photon/config/radeon.conf.

Load the configuration file and start the graphics driver by using the

-c option

of io-graphics:

io-graphics -c/usr/photon/config/radeon.conf

Here s the complete example:

GLOBAL

devices radeon

PLUGIN.photon

dllpath gri-photon.so

Here s a second example that shows how to use two different

devices for two displays:

devices banshee rage128

DEVICE.banshee

dllpath devg-banshee.so

pci_vendor_id 0x121a

pci_device_id 0x5

xres 1280

yres 1024

bitpp 32

DEVICE.rage128

dllpath devg-ati_rage128.so

pci_device_id 0x5050

region_x 1280

Troubleshooting your video card

Here are some things to check if

fails to detect

your video card or has problems:

The display is slightly misaligned on the monitor.

Each monitor behaves differently in certain resolutions and refresh

rates. If Photon seems to be slightly off center, you may need to

use the controls on your monitor to adjust the display to properly

align with the video mode. See your monitor s documentation for more

information.

The refresh rate isn t correct, the display isn t centered, or

in the worst case there s no display at all.

The timings set by drivers might not match the timings required by the

monitor or LCD.

You might have to edit /usr/photon/config/crtc-settings.

This file contains the timing information required by many drivers to set

the standard resolutions and refresh rates.

Many older drivers, as well as the generic drivers, don t use this file,

because they use the video BIOS to set the graphics mode.

If this file isn t present, the timings are calculated using a

generic timing formula GTF.

The GTF isn t completely accurate; you can manually edit the

crtc-settings file to set the exact timing for

the resolution and refresh rate.

At least the following drivers use this file:

devg-ati_rage128.so

devg-banshee.so

devg-chips.so

devg-flat.so

devg-i810.so

devg-i830.so

devg-igs5000.so

devg-mq200.so

devg-s3_savage.so

devg-sis630.so

devg-tnt.so

devg-tvia.so

devg-vmware.so

The modes detected are incorrect.

In some circumstances, a card s modes file doesn t list all the

modes that the documentation for the card says it should. Sometimes

there are variants of cards that are slightly different from what

we have had internally or what the documentation describes. You

can try manually setting up the card see

Manually setting up your video card,

above to see if that helps.

If it still fails, you may need to contact Technical Support.

The video mode changed to a mode that doesn t work.

If you ve changed the video mode to a mode that either your monitor

or graphics card doesn t support and that causes

Photon to be displayed improperly, you can revert to a previous mode, as

follows:

Restart the system and boot into safe mode don t start

Photon.

Clear the modes file. From the command line type: crttrap clear

Restart your system and try to detect your video card again.

The crttrap program hangs.

If crttrap hangs i.e. it doesn t start Photon

and leaves you with a black screen, try the following:

Ensure that your graphics card is supported.

Clear the modes file. From the command line type:

crttrap clear

If this fails, clear the modes file, but don t restart the system.

Then run crttrap manually from

the command line with the verbose option:

crttrap trap -VVVVV

If crttrap hangs again, you should be able to see

where it failed or what the last adapter it tried to detect was.

If you can determine the adapter that it failed to trap,

remove it from the trap list by editing

/etc/system/config/graphics-traplist.

The list that s generated comes from the enumerators, so you might have

to change the enumerators so they no longer detect

the card in your system, e.g. if you have two video cards in your

system.

You can find the enumerators in

/etc/system/enum/devices/graphics.

Once you ve removed the offending device from the traplist,

check to see if crttrap detects the card properly:

If crttrap still hangs, try setting the mode manually by

editing /etc/system/config/graphics-modes.

Create this file it if it doesn t already exist.

Also keep in mind that you can always run

the Vesa generic driver,

devg-vesabios.so.

Your graphics card isn t detected.

If you have a card that you expect to work with one of our accelerated

drivers, but you get only the generic drivers detected by crttrap,

you might have a chipset that s missing from the graphics enumerator.

Try the following:

Look in /etc/system/enum/devices/graphics for the

driver that you think you should run.

For example, if you re using a Radeon card, look for

devg-radeon.so.

Edit /etc/system/enum/devices/overrides and

add an entry for your card.

You need to specify the PCI Vendor and Device ID of the chipset, which

you can find by using the

pci

utility.

Add a line after the last known PCI ID and before

the line:

cfg fname, /etc/system/config/graphics-traplist

Save the file and reboot the system.

If you edited the enumerator correctly, there should be an entry in

/etc/system/config/graphics-traplist with the driver you

hope to run.

Run:

crttrap trap

The driver you want should now have been used and detected.

Sometimes this procedure works, other times it doesn t.

If it doesn t work, use one of the generic drivers or contact us about

adding support for your graphics chipset.

Even if this procedure works, we can t guarantee the driver will work as

expected.

If the device ID wasn t in the graphics enumerator, we haven t tried

that hardware and therefore can t guarantee it will work properly.

pci_attach_device

QNX Developer Support

Download the latest drivers for your PCI Device to keep your Computer up-to-date.

Attach a driver to a PCI device

Synopsis:

include

void pci_attach_device

void handle,

uint32_t flags,

uint16_t idx,

struct pci_dev_info info ;

Arguments:

handle

A handle that identifies the device.

The first time you call this function, set handle to

NULL.

This function returns a handle that you can use in a subsequent call

to allocate resources for the device.

flags

Flags that tell the PCI server how you want it to handle resources, which

resources to scan for, and which resources to allocate; see

Flags,

below.

idx

The index of the device: 0 for the first device, 1 for the second,

and so on.

info

A pointer to a

pci_dev_info

structure see below that specifies the

class code, vendor/device ID, or bus number and device/function number that

you want to scan for.

The function fills in this structure with information about the device.

Library:

libc

Use the -l c option to

qcc

to link against this library.

This library is usually included automatically.

Description:

The pci_attach_device function attaches a driver to a PCI

device.

You must successfully call

pci_attach

before calling any of the other PCI functions.

Typically drivers use this function to attach themselves to a PCI

device, so that other drivers can t attach to the same device.

If you specify the

PCI_SHARE flag see

below, then multiple drivers can attach to the same device.

The server can scan based on a class code, vendor/device ID, or bus number

and device/function number.

To control the server scanning, initialize the

appropriate fields of the info structure and set the

appropriate flags.

When you first attach to an uninitialized device, the PCI server assigns all

the I/O ports, memory and IRQs required for the device.

It also does the IRQ routing for you.

Once this has completed successfully, it fills in all these

values into your pci_dev_info structure to return these values

to your application.

When a driver attaches to a device, the PCI server allocates the

necessary resources for the device from procnto using the

rsrcdbmgr calls.

On X86 BIOS systems, these resources are normally allocated by the BIOS, but on

non-x86 systems, these resources have to be allocated from procnto.

You can detach the device by passing its handle to

pci_detach_device.

If you call

pci_detach,

any resources that pci_attach_device allocates are freed.

pci_dev_info structure

This function fills in a pci_dev_info structure that

describes an occurrence of a device.

The pci_attach_device function doesn t map any of the I/O

or memory regions into the process s address space.

The addresses returned in the pci_dev_info structure are all

physical addresses.

This structure has the following members:

uint16_t DeviceId

The device ID input/output.

For a list of supported device IDs, see.

uint16_t VendorId

The vendor ID input/output.

For a list of supported vendor IDs, see.

uint16_t SubsystemId

The subsystem ID output.

uint16_t SubsystemVendorId

The subsystem vendor ID output.

uint8_t BusNumber

The bus number input/output.

uint8_t DevFunc

The device/function number input/output.

uint8_t Revision

The device revision output.

uint32_t Class

The class code input/output.

For a list of class codes, see.

This field is an ORed combination of a class code and a subclass code e.g.

PCI_CLASS_DISPLAY PCI_SUBCLASS_DISPLAY_XGA.

uint32_t Irq

The interrupt number output.

uint64_t CpuIoTranslation

The CPU-to-PCI translation value

pci_addr cpu_addr - translation.

uint64_t CpuMemTranslation

The CPU-to-PCI memory translation

uint64_t CpuIsaTranslation

The CPU-to-ISA memory translation

uint64_t CpuBmstrTranslation

The translation from the CPU busmaster address to the PCI busmaster

address pci_addr cpu_addr translation.

uint64_t PciBaseAddress 6

The PCI base address array of six uint64_t items.

This function decodes bits 1 and 2 to see whether the register is 32 or

64 bits wide, hence the 64-bit values for the base registers.

uint64_t CpuBaseAddress 6

The CPU base address an array of six uint64_t items.

Some platforms translate addresses across PCI bridges, so that there s

one address on the PCI side of the bridge and another on the CPU side.

Under x86, the PciBaseAddress and CpuBaseAddress are

the same, but under other platforms, these will be different.

In your user application you should always use the CpuBaseAddress.

uint32_t BaseAddressSize 6

The size of the base address aperture into the board

an array of six uint32_t items.

uint64_t PciRom

The PCI ROM address.

uint64_t CpuRom

The CPU ROM address.

uint32_t RomSize

The size of the aperture into the board.

Flags

The flags parameter tells

the PCI server how resources are to be handled, which resources to scan for,

and which resources to allocate.

These bits control how resources are handled:

PCI_SHARE

Allow resources to be shared with other drivers. If this isn t set,

no other driver can attach to the device.

PCI_PERSIST

Resources persist after the device is detached.

The following bits ask the PCI server to scan for a device based on

the fields that you specified in the structure pointed to by info:

PCI_SEARCH_VEND

VendorID

PCI_SEARCH_VENDEV

DeviceId and VendorId

PCI_SEARCH_CLASS

Class

PCI_SEARCH_BUSDEV

BusNumber and DevFunc

These bits specify which members of the structure the server should

initialize:

PCI_INIT_IRQ

Irq

PCI_INIT_ROM

PciRom and CpuRom

PCI_INIT_BASE0 PCI_INIT_BASE5

The specified entries of the PciBaseAddress

and CpuBaseAddress arrays

PCI_INIT_ALL

All members except PciRom and CpuRom

The bits also include:

PCI_MASTER_ENABLE

Enable bus mastering on the device.

If you pass 0 for the flags, the default is PCI_SEARCH_VENDEV.

Testing and converting addresses

To facilitate the testing of addresses returned by the PCI server,

at least the following macros are defined in the header file:

PCI_IS_IO address

Test whether the address is an I/O address.

PCI_IS_MEM address

Test whether the address is a memory address.

PCI_IO_ADDR address

Convert the address returned by the PCI server to an I/O address.

PCI_MEM_ADDR address

Convert the address returned by the PCI server to a memory address.

PCI_ROM_ADDR address

Convert the address returned by the PCI server to a ROM address.

For example:

uint64_t port;

/ Test the address returned by the pci server /

if PCI_IS_IO addr

port PCI_IO_ADDR addr ;

Returns:

A handle to be used for other pci_ calls associated with

a handle, or NULL if an error occurs errno is set.

Errors:

EBUSY

An application has already attached to the device. If it s safe to share

the device, specify PCI_SHARE in the flags field.

EINVAL

The function couldn t attach a resource to the device.

ENODEV

This device wasn t found.

Examples:

Attach to and allocate all resources for the first occurrence of an Adaptec

2940 adapter:

int main void

int pidx;

void hdl;

int phdl;

struct pci_dev_info inf;

/ Connect to the PCI server /

phdl pci_attach 0 ;

if phdl -1

fprintf stderr, Unable to initialize PCI n ;

return EXIT_FAILURE;

/ Initialize the pci_dev_info structure /

memset inf, 0, sizeof inf ;

pidx 0;

inf.VendorId PCI_VENDOR_ID_ADAPTEC;

inf.DeviceId PCI_DEVICE_ID_ADAPTEC_2940F;

hdl pci_attach_device NULL, PCI_INIT_ALL, pidx, inf ;

if hdl NULL

fprintf stderr, Unable to locate adapter n ;

else

/ Do something to the adapter /

pci_detach_device hdl ;

/ Disconnect from the PCI server /

pci_detach phdl ;

return EXIT_SUCCESS;

Attach to the first occurrence of an Adapter 2940 adapter and allocate

resources in a second call:

void retval;

hdl pci_attach_device NULL, 0, pidx, inf ;

retval pci_attach_device hdl, PCI_INIT_ALL, pidx, inf ;

if retval NULL

fprintf stderr, Unable allocate resources n ;

Classification:

QNX Neutrino

Safety:

Cancellation point

Yes

Interrupt handler

Signal handler

Thread