>I may still be able to help you out, if you have access to a RISC PC - >I've decided to upgrade this box to StrongARM and RISC OS 3.7 since I can >do that within the time constraints I have to work with. When that >happens, the ARM710a board and RISC OS 3.6 ROMS will be surplus to my >requirements, and I'll happily donate them to the cause, if you can put >them to a good home. Thanks for the offer. But no, I don't have any RiscPC hardware at all. If you have time to run some tests, the program attached to this mail may help to find out what's going on. You need to find a simple invocation of grep, or whatever, that reliably fails. Then, run it like this: $ ssd grep "abc" /dev/null It will take an insanely long time to run, but eventually you will get a "ssd.log" file out which contains a trace of all the instructions executed by grep. Then we should be able to compare it to a similar trace on a machine where it works, to find out where the two diverge. Obviously for this to work the two machines need to be running exactly the same binaries, but that should be no big deal so long as you use the ones from the archive and don't compile your own. p.
Attachment:
ssd
Description: ssd
Attachment:
pgpIk6SDTw04v.pgp
Description: PGP signature