This is the mail archive of the ecos-discuss@sources.redhat.com mailing list for the eCos project.


Index Nav: [Date Index] [Subject Index] [Author Index] [Thread Index]
Message Nav: [Date Prev] [Date Next] [Thread Prev] [Thread Next]

RE: Enable EDB7xxx interrupts


> > > Is EINT1 brought out anywhere on the board? Have you scoped
> > > it? I take it PCCIRRx are registers that control whether or 
> > > not the pc card controller trips EINT1 on a card interrupt? 
> > > For PCIOSR you indicate that you can switch interrupts 
> > > between EINT1 and the fiq, have you tried setting up a 
> > > handler on the fiq to see if that is catching any of your 
> > > interrupts? Is EINT1 shared on that board? Is it possible 
> > > that something else is holding it in limbo? From the cirrus 
> > > web page, the cl-ps6700 can only be clocked at 13 or 18mHz, i 
> > > take it your bus speed is adjusted appropriately? It might 
> > > help if you included your definitions for PCCIRR1, PCPMR, 
> > > PCSICR, PCIOSR, and a brief snippet of the output from your 
> > > program. Also, out of convention, it's a bit more clear if 
> > > you use the pin definitions in hal_edb7xxx.h, such as 
> > > SYSCON2_PCMCIA1 instead of 0x020.
> > > 
> > > In general your interrupt usage looks ok, i would be more
> > > suspicious about the setup of the controller or the hardware 
> > > than eCos interrupt management.
> > 
> > Thanks for the reply
> > 
> > PCCIRR = PC Card Interrupt Reserved Register
> > 	These need to be set to all 1s in order for interrupts 
> to be enabled.
> > PCIOSR = PC Card Interrupt Output Select
> > 	This register selects which interrupts are output to what line.
> > Eg EINT1
> 
> Can you switch the line on these so they output to a 
> different interrupt line and listen on that line?

I just tried EXTFIQ, and it doesn't work either.

> > PCICR = PC Card Interrupt Clear Register
> 
> So presumably flipping a bit high in the interrupt clear 
> register clears a pc card interrupt.

If you mean setting it, yes.

> > I'm pretty sure EINT1 is not shared.  If it is, it's eCos 
> sharing it, 
> > and I don't know if it is or not.
> 
> By shared i mean does the board have EINT1 wired to two 
> physical interfaces, but i'm guessing cirrus probably didn't do that.

>From what others on the list have said, there are only two interrupts on
the EP7211 and everything are wired to them.  The Preliminary Data Sheet
says the same thing.  It says there are Seventeen interrupts sources
mapped to the IRQ line, and five sources mapped to the FIQ input.

> I dont have a 7211 laying around atm, but if the interrupt 
> lead off it is exposed i would scope that to make sure it's 
> actually generating an interrupt.

I tested the lines with a multi tester.  All the relevant lines
throughout the board are working.  The only problem is, is that the
CL-PS6700 controller is not pulling the lines low like it should when an
interrupt occurs.  It looks like the chip might be fried, but I'm not
sure yet.  I would think that if the controller is fully active like it
appears to be then there would be no need to do something to enable the
actual output of the interrupt lines, would there?  I mean the
interrupts appear to be working as far as the CL-PS6700 controller
registers are concerned.  If this is the case, wouldn't the be
outputting to the lines?


Index Nav: [Date Index] [Subject Index] [Author Index] [Thread Index]
Message Nav: [Date Prev] [Date Next] [Thread Prev] [Thread Next]