A request to realists from emulators
ZXNet echo conference «hardware.zx»
Hello, All
Very big. Anyone who has the real thing really needs to do a series of tests. Check it out
please here [http://zx.pk.ru/showthread.php?p=39944&posted=1#post39944].
We are waiting :o
Hello, Vladimir Kladov
Vla> Very large. Anyone who has real money really needs to do a series of tests.
Vla> Please look here
Vla> [http://zx.pk.ru/showthread.php?p=39944&posted=1#post39944]. We are waiting :o
Are we talking about this?:
=== : ===
boo_boo
I made a test on my knee for bit (hl)...
plz, whoever has real money at hand, run this thing there, and post the scr with
the chaos that formed on the screen.
boom compare
Attachments btst.zip (5.4 KB)
=== . ===
Result from "Pentagon 128".
File: btst.gif http://zx.pk.ru/attachment.php?attachmentid=2662
Hello, Wlodek
wow, thank you very much! decoding of this dregs - here
[http://zx.pk.ru/showthread.php?p=39969#post39969] .
Hello, Wlodek
Wlo> For an hour now she has been standing on aluop (a,b,c,d,e.....) and it’s not clear what she’s doing
Wlo> is there anything or hanging after the next turn on of the refrigerator
This is the longest test (by the way, it should fail). if it
throw it away, everything would go 4 times faster
Hello, goodboy
great, wow! everything is the same, except for LD (FFXX),A, which
sets the bits as described in the nocash doc! that means crystals
are still different. (and why not? for more than two decades, which
z80 is being released, something might have changed, but about undocumented “features” in
In this case, obviously no one will care). I’ll write a longer test in my spare time
and more informative in terms of the information provided.
PS Wlodek, what is written on your process?
Hello, boo_boo
I'm "afraid" that MemPtr may be influenced by int, even if it is not
triggered due to DI. For example, it can reset MemPtr or put it in
its "high" byte is the value of I.
Wlodek, keep it up! We will say a big inhuman THANK YOU!
Hello, boo_boo
Much more fun is the dregs from bo-bo in emulators. Pure 0, only real
spec was puffing up something, and in two squares it shows something different than
the rest.
Hello, Vladimir Kladov
I put here [http://zx.pk.ru/showthread.php?p=39984#post39984] trdos
version of zexall, will anyone undertake to run it in real life? (11 hours if without
turbo).
Hello, boo_boo
Test zexall takes about two and a half hours. Five lines have passed, in all
"OK" for now. The last one is "aluop a,nn".
Hello, Vladimir Kladov
2Vladimir Kladov:
But you can, most likely, you can auto-upload screenshots every 22
test lines. And make some kind of indication of the presence of program activity. Already
for an hour she stands on aluop (a,b,c,d,e.....) and it is not clear whether she is doing anything or
hangs after the next time the refrigerator is turned on.
Maybe you can upgrade the test in the next few hours and then I'll bubble it up
all night long, no questions asked?
In general, how relevant is this race in terms of the speed of testing?
? :)
Hello, SMT
allocation a,: CRC 1dfd7864
Expected: 06c7aa8e
aluop a, - goes...
Hello, Wlodek
Wlo> I'm joking, of course, especially since I got an ulcer and am no longer in
Wlo> able to get drunk on beer (doctors only allow vodka, and I don’t want it
Wlo> I can’t stand it). But I still wonder if mine was beneficial
Wlo> quarter-day vigil over the test at the Pentagon? :)
:o of course it brought it! we have come to the threshold of great discoveries and the creation of an ideal
emulation z80(tm)!!! oh, with such a pace of technological progress, the hour is uneven
Specky will remain only in museums and insanely rich collectors. and then
no one will know anything about these damned undocumented flags and
secret registers O__o
Hello, Wlodek
I didn't do the test. The trouble is that we need to check how it is
is consistent with the run of the person who ran it on his hardware.
Now I’ll take off the medals and see what happened (disconnected from the network
of course) :)
That is, we will do our test (and Bobo over there is already trying his best), and quickly, so that
Don't sit for hours. We needed this one to be long and boring. Great 10ks! Thanks in
anyway.
Hello, Wlodek
Here [http://zx.pk.ru/showthread.php?p=40052&posted=1#post40052] are my conclusions.
(Thanks CityAce, now at least I know how to use the url tag). Honestly: I'm glad
that I was wrong and that the test passed with a bang. But at the same time I'm sad because
now I have a test in front of my nose that shows that something is wrong in
Kingdom of Denmark :( i.e. I now know how to make the test
I ran it with an emulator (ignore the information about the hidden register). But at the same time
I'm sure this is the wrong way. And this means that we don’t have anything else about memptr
we know. And it seems like we don’t even know what we don’t know.
Hello, Wlodek
same
b
bc
c
d
de
e
- OK
Hello, Wlodek
I'm joking, of course, especially since I got an ulcer and am no longer able to
get drunk on beer (doctors only allow vodka, and I can’t stand it). But
I still wonder if my quarter-day vigil over the test was of any benefit
at the Pentagon? :)
Hello, Wlodek
The test worked!!!
01a, 02, 03.jpg
File: 01a.jpg http://zx.pk.ru/attachment.php?attachmentid=2679
File: 02.jpg http://zx.pk.ru/attachment.php?attachmentid=2680
File: 03.jpg http://zx.pk.ru/attachment.php?attachmentid=2681
Hello, Wlodek
The answers came so quickly that I no longer have time to copy them.
Therefore, the photo is from the screen.
File: 01.jpg http://zx.pk.ru/attachment.php?attachmentid=2678
Hello, Wlodek
h
hl
ix
iy
l
- OK
(hl) - error, I’ll write it off right now, messages ran quickly...
Hello, Wlodek
aluop a, - OK
aluop a,(+1)...
CRC: 52185ea5
exp: d3f2d74a
bit n,(+1)
CRC e1d4a896
exp: 83534ee1
bit n,
CRC: f554d742
exp: 5e020e98
cpd
CRC d06b9c15
exp: 134b622d
cpi
CRC 683e7b2a
exp: 2da42d19
- в процессе.
Hello, Wlodek
Vladimi Kladov, here is the Gold Star of the ZX Spectrum Hero, two 400 ml
"Hemodez" ampoules, one 400-ml "Chlosol" ampoule, one disposable
IV and eternal honor and respect ;) :) :) ;) :) . Shchyutka :) :) ;) .
Hello, Wlodek
(hl)
CRC: 1ca0ece9
exp: 28299ece
SP - OK
(ix,iy)+1)-1
CRC ff601665
exp 0b95a8ea
ixh
ixl
iyh
iyl
ld , (nnnn)
ld hl, (nnnn) - эти все OK.
Hello, Wlodek
- OK
Hello, Vladimir Kladov
I didn’t understand something: will there be a medal?? ;) :) :) :) :)
Hello SMT
SMT> except that the result is not specified ,
SMT> everything matches the old data
Am I missing something?? Well then, do a test with automatic uploading of screenshots. I don't
able to sit in front of a screen for half a day ;) .
Hello, Wlodek
except that the result is not specified, everything matches
with old data
Hello, CHRV
CHR> Specifically, I have the following zoo:
CHR> T34ВМ1
CHR> 1858ВМ1
CHR> 1858ВМ3
CHR> Z80B
CHR> Z0840008
CHR> Z084C010
wow!!! *licks*
Hello, boo_boo
Well, I wrote another test, a little heavier. please run it in real life! will work
quickly, after launching all sorts of letters will run across the screen, you DO NOT need to write them down
-- when the test passes, it will ask for the file name and send the log there as text.
this is the log that is needed...
thank you! :)
PS yes, here’s another thing - it’s better to run in 48 mode; when testing IN/OUT it gets into
all sorts of ports, I didn’t really keep track of which ones
File: btest2.zip http://zx.pk.ru/attachment.php?attachmentid=2685
Hello, boo_boo
boo> here, I wrote another test, a bit more weighty. please run it in real life!
boo> will work quickly, after launching all sorts of letters will run across the screen, them
boo> DO NOT write down - when the test passes, it will ask for the file name, and
boo> will send the log there as text. this is the log that is needed...
boo> thanks! :)
Yeah, collect all the tests but so that they automatically save the information!
I don’t mind testing everything on a bunch of different processors. In particular I have
next zoo:
T34VM1
1858VM1
1858VM3
Z80B
Z0840008
Z084C010
Hello, CHRV
Cool! Boo-boo, please post the source. Or then inject yourself. Especially on IM2
interesting. Although I don’t understand everything with CPD either.
Hello, Vladimir Kladov
great!!! the picture is starting to become clearer... but this is not the end :rolleyes:
hmmmm [http://zx.pk.ru/showthread.php?p=40182#post40182]
Hello, boo_boo
Here I am again with the test ;)
the amount of information displayed (and, accordingly, scrolls) has increased noticeably, but if in
At the beginning, answer the question with the “2” button; you don’t have to press enter.
thank you :)
File: btest4.zip http://zx.pk.ru/attachment.php?attachmentid=2705
Hello, boo_boo
Would it be possible to immediately add LDI/LDD/CPI/CPD to the test with BC=FFFF, FFFE, FF00
Hello, Vladimir Kladov
Vla> Or you could immediately add LDI/LDD/CPI/CPD to the test with BC=FFFF, FFFE,
Vla> FF00
possible, perhaps. I'll add if this test doesn't clear things up..
Hello, boo_boo
boo> is possible, perhaps. I'll add if this test doesn't clear things up..
SORRY guys, I'll be able to run tests on Friday :(
So it’s better to plug everything in at once to watch it right away!
Hello, Vladimir Kladov
Vla> My point is that I don’t like the word “resets”. I think that
Vla> no special dumping is ever done, data is always taken
Vla> from somewhere. In this regard, I would still think about what kind of reset this is
Vla> in the command LD (xxxx), A. However, I don’t know what kind of register there might be
Vla> be a "reset" source. But there is definitely something there. Maybe PC, or F.
maybe so. In this regard, block I/O instructions are especially interesting, since
in the last test they "reset" memptr for all processors, not just for clones
(if Wlodek has a clone, of course)
Hello, CHRV
Gentlemen, who are really interested in the topic and have rare processors,
in particular:
- T36VM1 (Angstrom)
- U80A MME (German DR)
- ST Z80A (ST company)
- Z80B,H (Zilog)
- some others
Give them or send them to me by mail (see website), in return I will provide you
Z0840008 (Zilog) and NedoPC logs. So that there is interest.
Hello, boo_boo
boo> ok :rolleyes: now everything seems to be plugged in, and even a little extra
boo> (in addition to 2 bits of memptr, flags are also output after the one being tested
boo> instructions, and these flags are added to the sum, and for IN** by mask
boo> - a step towards CPUID density ;) )
Results.
1) Z0840008
2) T34VM1
3) KR1858VM3
4) Z084C0010
5) KR1858VM1
File: BTEST5I.ZIP http://zx.pk.ru/attachment.php?attachmentid=2728
Hello, CHRV
From Wlodek with the fifth greeting.
File: 1wlodek5.zip http://zx.pk.ru/attachment.php?attachmentid=2729
Hello, Wlodek
Wlo> From Wlodek with the fifth greeting.
:)
Hello, boo_boo
So I did it on the ZX Spectrum+. There are two problems (more precisely, one, the other
question). The output is recorded in a wav file, I've been struggling for an hour trying to translate it
it is either in tap or in tzx. Any suggestions on this matter? Second - there are two in the test
options, non-stop and normal. I ran both, hence two files. Which of
do you need them?
There are still problems with the Sprinter - it does not want to read discs. But I’ll figure this out somehow.
Hello, boo_boo
Well, I've finally surpassed it. Both files are in the archive (see previous post). TZX format.
File: btest5.zip http://zx.pk.ru/attachment.php?attachmentid=2732
Hello icebear
ice> So I did it on the ZX Spectrum+. There are two problems (or rather one,
ice> another question). The output is written to a wav file, I've been working on it for an hour
ice> to convert it to either tap or tzx. There is something about this
ice> suggestions? Secondly, the test has two options, non-stop and normal. I drove away
ice> both, hence two files. Which one is needed?
about Vawa - http://www.ramsoft.bbk.org/maketzx.html. but the test is the same
for both options, it’s just that “scroll?” appears in the first one, but not in the second one
:rolleyes:
Hello, Vladimir Kladov
again a test designed to crown the research with either a victorious point or
another ellipsis ;)
block commands in different variations at different addresses.
it goes to 48k itself, first asks for a name, writes 8 files (“Name_number”).
plz, drive away! :rolleyes:
For some reason, the volume (and running time) of tests is increasing and increasing...
maybe it’s time to attach a runaway line with music to them to combine
useful with pleasant? O__o
File: bltst.zip http://zx.pk.ru/attachment.php?attachmentid=2756
Hello, boo_boo
boo> again a test designed to crown the research with either a victorious point
boo> or another ellipsis ;)
boo> block commands in different variations at different addresses.
boo> goes to 48k by itself, first asks for a name, writes 8 files
boo> ("Name_number").
boo> plz, banish! :rolleyes:
boo>
boo> For some reason, the volume (and running time) of tests is increasing and
boo> is increasing... maybe it's time for the escaped line with music
boo> attach to combine business with pleasure? O__o
I'll drive in the evening!
And this is a normal CPUid utility with music and beauty :).
For debugging, including emulators.
By the way, you can include there the calculation of the number of clock cycles, the current frequency of the processor, and
much more!
It would be a very useful tool.
Hello SMT
And please run my DAA test at the same time. At least on 2 computers, no more
necessary.
Hello, Vladimir Kladov
Тупил я долго и упорно, а все из-за некоего Мака Бастера (или того, кто за него
этот текст вбивал):
Hobeta
[From Mac Buster]
Hobeta files (which use the extension .$?) is simply an image of a TR-DOS file,
rather than a complete disk; it is just the file with a 17-byte header:
Offset Name Size Description
0x00 FileName 0x08 TR-DOS file name
0x08 FileType 0x01 TR-DOS file type
0x09 StartAdr 0x02 start address of file
0x0A FlLength 0x02 length of file (in bytes)
0x0C FileSize 0x02 size of file (in sectors)
0x0E HdrCRC16 0x02 Control checksum of the 15 byte
header (not sector data!)
A Hobeta file cannot be smaller than 256+17 bytes, or larger than 65280+17
bytes (as a TR-DOS file can have be a maximum of 255 sectors long). The file
must be a 17+a multiple of 256 bytes long.
Конечно, сам виноват, но докам даже полу-официальным как-то привык верить.
Проверить, что 9+2 не могет никак быть = 10 просто в голову не пришло. В итоге
прога брала со смещения на 1 меньше чем надо, и поэтому первый результ в блоке
и не мог сойтись.
Так что неправильные результаты дает пока только спектакулятор. Хочетя увидеть
есть ли другие результаты чем у Влодека на других машинках. А прогу дляI’ll update the comparison of results in a moment in the one below/above (it seems like yesterday)
post.
But while I was debugging, I found 2 bugs, I’ll fix them now :)
Hello, Vladimir Kladov
Vla> And please run my DAA test at the same time. At least on 2 computers
Vla> no more is needed.
Yes of course :)
Hello, Vladimir Kladov
bltest and DAA test results.
The following processes were used:
1) Z0840008
2) T34VM1
3) 1858VM1
4) 1858VM3
5) Z084C00010
You'll figure it out in short :).
I almost forgot - mode 48, without turbo.
File: bltsti.ZIP http://zx.pk.ru/attachment.php?attachmentid=2765
Hello, boo_boo
boo> again a test designed to crown the research with either a victorious point
boo> or another ellipsis ;)
boo> block commands in different variations at different addresses.
boo> goes to 48k by itself, first asks for a name, writes 8 files
boo> ("Name_number").
boo> plz, banish!
boo>
Result from Wlodek.
File: w-bl0603.zip http://zx.pk.ru/attachment.php?attachmentid=2776
Hello, Vladimir Kladov
Vla> I haven’t looked at the new tests yet, just with an eye. First number above
Vla> test block - what does it mean? F7FF for example.
address of the instruction/chain being tested.
the instructions hide such a swamp ;)
btw, no one has driven t80 to fpga? I'd like to check it out...
Hello, CHRV
DAA results:
CHRV 1) Z0840008 = Wlodek
CHRV 2) T34BM1 did not match on two lines:
AFsrc=3FFF, CHRV2: AFout=02DB Wlodek: AFout=D98B
AFsrc=BFFF, CHRV2: AFout=02DB Wlodek: AFout=2531
CHRV 3) 1858VM1 did not match on one line:
AFsrc=BFFF, CHRV3: AFout=02DB Wlodek: AFout=2531
CHRV 4) 1858VM3 did not coincide with Wlodek’s on 1 line:
AFsrc=F7AF, CHRV4: AFout=0001 Wlodek: AFout=5701
CHRV 5) Z084C00010 = Wlodek
Here is CPUID in its “pure” form, at least Soviet clones
recognized with a bang :) So, CHRV, get some more percent, and how will it be
run it, the DAA result is very indicative! (And I'll have to increase
number of different stones in the emulator configuration).
Hello, Wlodek
I haven’t looked at the new tests yet, just glancing at them. The first number above the test block is
what does it mean? F7FF for example. But then I’ll definitely take a closer look, I took it for myself
looking and thinking. (If CPD:CPD acts as CPDR and CPIR:CPD as CPIR,
then the conclusion is that most likely the work in them goes through different registers, and they
are stored in memptr, which serves as a latch buffer for them, before
sent to the bus or to an internal register. The fact that the registers are different is understandable.
One is an incremental counter, the other is a decremental counter. It's still possible there is
separate adder - and there may be a separate subtractor although this may be
summation with a negative, it can be obtained by inversion and decrement - with two
input and one output register. That is then we need to look at the mutual influence
such groups: CPI:ADD HL,DE:CPI for example). I think life would be simpler if
If only the developers of the stone would hide this mempter out of sight and not stick it out
out. They felt the urge to do such a disgusting thing, now we are torn apart for
them this mess......
I took a look at DAA (I posted the results, the CPUID looks good on DAA), and I’ll get distracted by
holidays for other matters.
Hello, boo_boo
hi, it's me again with a test :rolleyes:
testing the hypothesis about the existence of 2 memptrs, further investigations with
CP* and OUT*..
I think this time there will be enough results for a couple of processors, one of which
zilog-ovskiy..
:)
File: bltst2.zip http://zx.pk.ru/attachment.php?attachmentid=2908
Hello, Wlodek
Here’s another test, I really hope it’s the final one. it would be on several different CPUs
drive away...
File: mptrtst.zip http://zx.pk.ru/attachment.php?attachmentid=2953
Hello, boo_boo
boo> here's another test, I really hope it's the final one. it would be on several
boo> drive different cpus...
Stas, I'll take a look one of these days.
If you're going to change anything, change it now :)
Hello, boo_boo
boo> here's another test, I really hope it's the final one. it would be on several
boo> drive different cpus...
Greetings on the topic mptrtst!
I couldn't do it on the Pentagon in either 128K or 48K.
At the end, it prints thanks and hangs itself, having previously initialized
CDOS modem and picking up the phone on it.
By the way, about "Scroll?" Is it necessary to answer manually in these tests?
to this question? Maybe you can just automate the scrolling of text on
screen, constantly sending the number 255 to cell 23692 in the test program itself? Everything
anyway, the one who tests is unlikely to read the output ;) .
Hello, Wlodek
Wlo> Greetings on the topic mptrtst!
Wlo> I couldn't run it on the Pentagon in either 128K or 48K.
Wlo> At the end he displays gratitude and hangs himself, preliminary
Wlo> initializing
Wlo> CDOS modem and picking up the handset on it.
oh those ports! sinful, I thought, since I cut off 7ffd, any dirty tricks can be done with
playing with ports :rolleyes:
fixed everything on xxFE.
Wlo> By the way, about "Scroll?" Is it necessary to answer manually in these tests
Wlo> answer to this question? Maybe you can just automate text scrolling
Wlo> on the screen, constantly sending the number 255 in cell 23692
Wlo> test program? Anyway, the one who tests is unlikely
Wlo> is read into the output ;) .
item 2) in the initial menu does this: rolleyes: apparently I named it
unsuccessful, fixed it to "auto scroll"
here it is :)
File: mptrtst.zip http://zx.pk.ru/attachment.php?attachmentid=2971
Hello, Wlodek
I ran it on the processors (I write everything that is written on them):
1. SGS Z8400B81 (Z80BCPU - 28741)
2. Zilog Z0840004PSC (Z80 CPU - 8916)
3. Zilog Z0840004PSC (Z80 CPU - 9324)
File: log_2803.zip http://zx.pk.ru/attachment.php?attachmentid=2977
Hello, boo_boo
boo> Great, thanks! everything is as I thought!
boo> CHRV, I’m waiting for more results from you (just, please, take the last one
boo> version of the test - I changed it a little to make sure a couple of things
boo> check) :) , and when I get it, I’ll write a detailed poem about memptr
boo> :rolleyes:
Actually the poem to the studio :).
Or better yet, you can post the article on my website!
The test involved Turbo2+ and:
1) Z0840008
2) Z84C0010
3) KR1858VM3
4) KR1858VM1
5) T34VM1
File: MPRTST_R.zip http://zx.pk.ru/attachment.php?attachmentid=2983
Hello, molodcov_alex
Great, thanks! everything is just as I thought!
CHRV, I’m still waiting for results from you (just, plz, take the latest version of the test
- I changed it a little, just to check a couple of things for sure) :), and how
I'll wait until I write a detailed poem about memptr: rolleyes:
Hello, CHRV
CHR> Actually the poem to the studio :).
CHR> Or better yet, you can post an article on my website!
I don’t understand whether it was a poem or an article :rolleyes:
mulberry! [http://zx.pk.ru/showthread.php?p=43688#post43688]
Hello, boo_boo
Gone to Fido!
Hello, Vladimir Kladov
Vla> Run the test only on the Pentagon - Scorpios are not suitable. (If in
Vla> ATM is fully compatible with the Pentagon, drive in this mode). Ha
Vla> proprietary models the test is not calculated at all.
What does the test do first?
The FD port on the modified Turbo2+ works the same as on the Pentagon.
Hello, Vladimir Kladov
Vla> It’s my turn to pester the real people with requests. This time only
Vla> Pentagon owners, please help and drive away this simple TRD
Vla> (I have attached other files for completeness).
Vla>
Vla> the test prompts you to enter a correction number from before each pair of checks
Vla> 0 to 30.
Vla> If the correction is entered correctly, then the green curb should stand and
Vla> don't move at all!
Vla> And then I wonder how the border will behave after pressing the next one
Vla> buttons. If it remains in place, then IM0=13 clock cycles, the same as IM1. A
Vla> if it goes (up), then = 12 clock cycles.
Vla>
Vla> Also of interest is the magnitude of the correction at which it is achieved
Vla> curb stop.
Vla>
IM1/IM0 test on a real Pentagon-128 in 48K mode.
In both cases, the strip is stationary at coefficient = 14.
Hello, Vladimir Kladov
By the way, if 13 instead of 14, then the test floats up? And then suddenly there is more
"rounding" of clock cycles in a frame due to some additional waits - I'm in
I don’t understand much about this, although it was explained to me a long time ago that this doesn’t happen (and
suddenly).
If there is a branded one, then tell me which one: it will take me a long time to adjust to everyone at once and
tedious, it’s better to know in advance which model to adjust the test to. He doesn't
self-adjusting, it’s hard for me to draw, I’ve never done anything like that
similar. It’s easier to adjust separately for each individual model. That's why
asked the Pentagon, it’s easiest to adjust to it.
And I also suspect that the issue may be the end of VK1 in Wlodek’s process. So,
CHRV, if you run at least one Pentagon with a bourgeois stone on board,
it will be a big help even if the test shows the same thing. (Extra
confirmation is not superfluous, suddenly IM0 in “real” Z80 costs 12 clock cycles).
Hello, Wlodek
Wonderful result. Thank you, Wlodek. The test shows that the duration
interrupts in IM0 mode exactly = duration in IM1 mode. Remains
repeat the “feat” for a branded spec. It’s hard to believe that the author
The Spectator simply believed this document (the author of which is also the author of the emulator
ZX32 for Windows, Greek Karpatzianis), and didn’t bother to check that it’s actually
that's true. For now, I remain in the position that for branded models this may suddenly
turn out to be true.
14 is great. This means that it’s only in my emulator that the roof goes off at about 4 clock cycles.
Okay, now I have a test where I can track this matter,
catch and punish.
2CHRV: the peculiarity of the test is that it works in enabled interrupts, but
Doesn't use the HALN command. Instead ensures that the loop lasts exactly 71680
bars (subject to a small amendment - just in case), including those bars
which are spent processing RST 38, which fires on every frame.
And so that RST 38 works in the same number of clock cycles, the test is at the beginning of each
loop restores system variables to the same state (0)
keyboard and frame counter.
Those. if IM0 had 12 clock cycles, then as a result of replacing the IM1 instruction with IM0
the entire cycle would decrease in time by exactly 1 beat, and the curb would begin graduallymove upward (this is noticeable only after a few seconds, and the expected
the effect can be observed by reducing the “correction” by exactly 1 before starting
another pair of tests, i.e. by entering 13).
By the way, the source code for the ASMA test is attached. If there are branded models, I can
try to match it to the company. It’s easier to customize in emulsion (and customize all sorts of things)
easier than sitting and calculating how many clock cycles each command takes).
Hello, Vladimir Kladov
Vla> And I also suspect that the issue may be the end of VK1 in the process
Vla> Wlodek'a. So, CHRV, if you run it on at least one Pentagon with
Vla> bourgeois stone on board, it will be of great help, even if
Vla> test will show the same thing. (An extra confirmation is not superfluous, suddenly
Vla> IM0 in “real” Z80 costs 12 clock cycles).
I'll try the Penatgon 1024SL, but I won't say exactly when, maybe on the weekend!
Hello, CHRV
CHR> I'll try it on Penatgon 1024SL, but I won't say exactly when, maybe in
CHR> weekend!
Hmm. It’s okay, I’ll wait another week, no problem. At least two. But the question is
fundamental, requires clarification.