Debug int vectors mirrored on 256k boundaries bug?

Peter Jakacki

During my testing I found that the debug vectors at $0F.FF80 are also mirrored on 256k boundaries.

$3.FF80 $80 DUMPL
03.FF80: 0000.0000 0000.0000 0000.0000 0000.0000    ................
03.FF90: 0000.0000 0000.0000 0000.0000 0000.0000    ................
03.FFA0: 0000.0000 0000.0000 0000.0000 0000.0000    ................
03.FFB0: 0000.0000 0000.0000 0000.0000 0000.0000    ................
03.FFC0: FABB.FFFF FABB.FFFF FABB.FFFF FABB.FFFF    ................
03.FFD0: FABB.FFFF FABB.FFFF FABB.FFFF FABB.FFFF    ................
03.FFE0: FABB.FFFF FABB.FFFF FABB.FFFF FABB.FFFF    ................
03.FFF0: FABB.FFFF FABB.FFFF FABB.FFFF FABB.FFFF    ................ ok

I know so because I was wiping large chunks of memory to test ERASE and FILL and it kept crashing!

Maybe it's just a DE2-115 thing. Not enough RAM?

cgracey

Even with the full 512KB, it still wraps twice.

Peter Jakacki

This is a map of 1M of memory in 1k blocks so that any non-zero block is reported.

$400 MAP
 0000.0000 0000.0400 0000.0800 0000.0C00 0000.1000 0000.1400 0000.1800 0000.1C00
 0000.BC00 0000.C000 0000.C400 0003.FC00 0004.0000 0004.0400 0004.0800 0004.0C00
 0004.1000 0004.1400 0004.1800 0004.1C00 0004.BC00 0004.C000 0004.C400 0007.FC00
 0008.0000 0008.0400 0008.0800 0008.0C00 0008.1000 0008.1400 0008.1800 0008.1C00
 0008.BC00 0008.C000 0008.C400 000B.FC00 000C.0000 000C.0400 000C.0800 000C.0C00
 000C.1000 000C.1400 000C.1800 000C.1C00 000C.BC00 000C.C000 000C.C400 000F.FC00 ok

After the first 64k where I have a lot of stuff it jumps up to 3.FC00 and now I realize that I only have 256k of memory with the vectors at the top of that. Is that right?

cgracey

Peter Jakacki wrote: »

This is a map of 1M of memory in 1k blocks so that any non-zero block is reported.

$400 MAP
 0000.0000 0000.0400 0000.0800 0000.0C00 0000.1000 0000.1400 0000.1800 0000.1C00
 0000.BC00 0000.C000 0000.C400 0003.FC00 0004.0000 0004.0400 0004.0800 0004.0C00
 0004.1000 0004.1400 0004.1800 0004.1C00 0004.BC00 0004.C000 0004.C400 0007.FC00
 0008.0000 0008.0400 0008.0800 0008.0C00 0008.1000 0008.1400 0008.1800 0008.1C00
 0008.BC00 0008.C000 0008.C400 000B.FC00 000C.0000 000C.0400 000C.0800 000C.0C00
 000C.1000 000C.1400 000C.1800 000C.1C00 000C.BC00 000C.C000 000C.C400 000F.FC00 ok

After the first 64k where I have a lot of stuff it jumps up to 3.FC00 and now I realize that I only have 256k of memory with the vectors at the top of that. Is that right?

The DE2-115 can only provide 256KB, so that sounds right.

Peter Jakacki

In testing multiple cogs with Tachyon code I found only 10 blinking plus my main console cogs. So that's 11 cogs which I also confirmed with the all_cogs_blink.

So the DE2-115 runs 11 cogs with 256k memory.