The use of IRQ 2 in an AT system deserves a bit of explanation. An AT uses IRQ 2 right on the motherboard, which means that the expansion bus pin for IRQ 2 is now empty. Instead of leaving this pin unused, IRQ 9 from the AT extended slot is wired to the pin previously occupied by IRQ 2.
In other words, IRQ 9 is being redirected to IRQ 2. Any AT expansion device set to use IRQ 2 is actually using IRQ 9. Of course, the vector interrupt table is adjusted to compensate for this slight of hand. After an interrupt is triggered, an interrupt-handling routine saves the current CPU register states to a small area of memory (called the
stack), then directs the CPU to the interrupt vector table. The interrupt vector table is a list of program locations that correspond to each interrupt. When an interrupt occurs, the CPU will jump to the interrupt handler routine at the location specified in the interrupt vector table and execute the routine.
Inmost cases, the interrupt handler is a device driver associated with the board generating the interrupt. For example, an IRQ from a network card will likely calla network device driver to operate the card. Fora hard-disk controller, an IRQ calls the BIOS ROM code that operates the drive. When the handling routine is finished, the CPU’s original register contents
are popped from the stack, and the CPU picks up from where it left off without interruption. As a technician, it is not vital that you understand precisely how interrupts are initialized and enabled, but you should know the basic terminology. The term
assigned simply means that a device is set to produce a particular IRQ signal. For example, atypical hard drive controller board is assigned to IRQ 14. Assignments are usually made with one or
more jumpers or DIP switches, or are configured automatically through the use of Plugand - Play
(PnP). Next, interrupts can be selectively enabled or disabled under software control. An enabled interrupt is an interrupt where the PIC has been programmed to pass on an IRQ to the CPU. Just because an interrupt is enabled does not mean that there 258 CONFLICT TROUBLESHOOTING are any devices assigned to it. Finally, an active interrupt is a line where real IRQs are being generated. Notice that
active does not mean assigned or enabled. Interrupts are an effective and reliable means of signaling the CPU,
but the conventional ISA bus architecture—used in virtually all PCs—does not provide a means of determining which slot contains the board that called the interrupt. As a result, interrupts cannot be shared by multiple devices. In other words, no two devices can be actively generating interrupt requests on the same IRQ line at the same time. If more than one device is assigned to the same interrupt line, a hardware conflict can occur.
Inmost circumstances, a conflict will prevent the newly installed board (or other previously installed boards) from working. In some cases, a hardware conflict can hangup the entire
Recognizing and Dealing with Conflicts Fortunately, conflicts are almost always the result of a PC upgrade gone awry. Thus, a technician can be alerted to the possibility of a system conflict by applying the Last Upgrade rule. The
rule consists of three parts 1 Apiece of hardware and/or software has been added to the system very recently.
2 The trouble occurred after apiece of hardware and/or software was added to the system.
3 The system was working fine before the hardware and/or software was added.
If all three of these commonsense factors are true, chances are very good that you are faced with a hardware or software conflict. Unlike most other types of PC problems, which tend to be specific to the faulty sub-assembly, conflicts usually manifest themselves as much more general and perplexing problems. The following symptoms are typical of serious hardware
or software conflicts - The system locks up during initialization.
- The system locks up during a particular application.
- The system locks up when a particular device (e.g., a TWAIN scanner) is used.
- The system locks up randomly or without warning regardless of the application.
- The system might not crash, but the device that was added might not function (even though it seems properly configured. Devices that were in the system previously might still work correctly.
- The system might not crash, but a device or application that was working previously no longer seems to function. The newly added device (and accompanying software) might notwork properly. What makes these problems so generic is that the severity and frequency of a fault, as well as the point
at which the fault occurs, depends on such factors as the particular deb
Share with your friends: