CD1284 Datasheet, PDF (102/176 Pages) Intel Corporation – IEEE 1284-Compatible Parallel Interface Controller with Two High-Speed Asynchronous Serial Ports

CD1284 — IEEE 1284-Compatible Parallel Interface Controller
6.5.2
6.6
102
break;
}
outportb(PFCR, inportb(PFCR & 0xEF);/* clear IntEn (first step of ‘toggle’
operation */
outportb(PFCR, inportb(PFCR | 0x10);/* set IntEn (second step of ‘toggle’
operation */
outportb(CAR, save_car);
/* restore original CAR (if desired) */
return(0);
}
Hardware-Activated Service Examples
Hardware-activated context switching is nearly identical to the serial case; during the service
acknowledge cycle, the SVCACKP* input is active and the CD1284 drives the parallel channel
vector on the data bus. At the same time, the MPU pushes the current state of the device on the
context stack and sets the context for channel 0. The vector comes from the PIVR, which is a
reflection of the LIVR. The vector supplied indicates the source of the request in the IT2-IT0 bits.
There is no equivalent to the Interrupting Channel register (TICR, RICR, MICR) since, by
definition, the interrupt is from channel 0. Once the context switch occurs, the CPU can proceed to
service the source of the request.
The CPU must decode the ITx bits to determine the blocks that require service. Each section of the
parallel channel has an Interrupt Status register to indicate what conditions, if any, in that block
require service. These are the PFSR in the data path and the PCISR in the channel control state
machine.
At the end of the service routine, the CPU must perform the same dummy write operation to the
EOSRR as for the serial channels. This informs the device that the parallel service is complete.
The write operation to the EOSRR generates a high-priority interrupt to the MPU to cause it to pop
the context stack and restores the device environment to what it was at the start of the interrupt
service.
Baud Rate Derivation
/* This is a simple code example which shows a way to derive the proper values for
the RCOR/TCOR and RBPR/TBPR register pairs for any baud rate. Routine is called with
the desired baud rate and master clock; global variables cor and bpr are set by the
routine. */
int
brp, cor;
compute_baud(clock, baud_rate)
double
clock;
double
baud_rate;
{
double
int i;
cor_values[ ] = {8.0, 32.0, 128.0, 512.0, 2048.0, -1.0};
for ( i = 0; cor_values[i] != -1; i++ )
{
brp = (int) ((( clock / baud_rate) / cor_values[i]) + 0.5);
if (brp < 0xFF)
{
cor = i;
bpr = brp;
break;
}
}
Datasheet