fun bug
ZXNet echo conference «real.speccy»
I caught an interesting bug:
ld bc,#7ffd:ld a,#91:out (c),a
ld hl,#c000:ld de,#7000:ld bc,#4000:ldir
At the same time, in bank #92 (yes, exactly in #92!!) there was some data, but
after ldir they crashed. Along the way, my hardware is malfunctioning - it does not have time to set the address corresponding to bit 7 of port 7ffd to zero. Instead of bank 2 at
addresses #8000-#bfff there is bank #92. The most interesting thing is that, besides my own program, the glitch did not manifest itself anywhere else (possibly due to the fact that
there is not so much software that uses 512kb on the Pentagon)
ld bc,#7ffd:ld a,#91:out (c),a
ld hl,#c000:ld de,#7000:ld bc,#4000:ldir
At the same time, in bank #92 (yes, exactly in #92!!) there was some data, but
after ldir they crashed. Along the way, my hardware is malfunctioning - it does not have time to set the address corresponding to bit 7 of port 7ffd to zero. Instead of bank 2 at
addresses #8000-#bfff there is bank #92. The most interesting thing is that, besides my own program, the glitch did not manifest itself anywhere else (possibly due to the fact that
there is not so much software that uses 512kb on the Pentagon)
JTN/4D
Press RESET immediately, Vassili!
12 Apr 02 19:16, Vassili Klimov wrote to All:
VK> I caught an interesting bug:
VK> ld bc,#7ffd:ld a,#91:out (c),a
VK> ld hl,#c000:ld de,#7000:ld bc,#4000:ldir
VK> At the same time, in bank #92 (yes, exactly in #92 !!) there was some
VK> data, but after ldir they crashed. Along the way, my hardware is glitching -
VK> does not have time to set the address corresponding to bit 7 of port
VK> 7ffd to zero.
It’s unlikely to have time. More likely, the wire is shorted somewhere with another discharge.
VK>> I caught an interesting bug:
VK>> ld bc,#7ffd:ld a,#91:out (c),a
VK>> ld hl,#c000:ld de,#7000:ld bc,#4000:ldir
VK>> At the same time, in bank #92 (yes, exactly in #92!!) there were some
VK>> data, but after ldir they crashed. Along the way, my hardware is glitching -
VK>> does not have time to set the address corresponding to bit 7 of port
VK>> 7ffd to zero.
VK>> ld bc,#7ffd:ld a,#91:out (c),a
VK>> ld hl,#c000:ld de,#7000:ld bc,#4000:ldir
VK>> At the same time, in bank #92 (yes, exactly in #92!!) there were some
VK>> data, but after ldir they crashed. Along the way, my hardware is glitching -
VK>> does not have time to set the address corresponding to bit 7 of port
VK>> 7ffd to zero.
I see you as if in reality, Vassili!
...14 Apr 02 20:46:51, Vassili Klimov -> Kirill Frolov, theme... VK>> hardware - does not have time to set to zero the address corresponding to
...
VK>> bit 7 of port 7ffd.
VK>