... Of course there is - you use a DC mmX'0' or DC mmXL(nn)'0', but the length expression must be valid. E.g., ORIGIN DS 0D ORG TEXTSTRT+csectname CODEBASE DS...
4786
Tony Harminc
tzha1
Jul 26, 2010 4:25 pm
... You keep referring to "the timer", and I think by this you mean the Interval (location 80) timer. I suggest that you handle this timer only to the extent...
4787
mfnoel
Jul 26, 2010 4:40 pm
I've posted the 5th (and final) in a series of KICKS KooKbooK recipes relating to my successful project to demonstrate over 250 users running at once on...
4788
kerravon86
Jul 26, 2010 5:44 pm
... Thanks, that was close enough to what I was after: CODESTRT EQU 1024 Start of our real code (Dave Wade - as you suggested) CSECT (name of my...
4789
Gerhard Postpischil
gerhard_post...
Jul 26, 2010 5:54 pm
... That's too much work. Starting with low storage that's zero, the timer interrupt will kick in when it goes negative. At that point it has wrapped around,...
4790
kerravon86
Jul 26, 2010 7:35 pm
... PDOS and NetBSD are two separate projects, other than the fact that I expect that experience of PDOS should provide assistance and code for use by NetBSD. ...
4791
Tony Harminc
tzha1
Jul 26, 2010 7:54 pm
... Sure. But by the same token, you can be even simpler: you never have to enable for external interrupts at all, and so don't need an external new PSW that...
4792
PeterH
peterh5322
Jul 26, 2010 8:12 pm
... FWIW, see the most recent edits to the PL/S topic in Wikipedia. I added those, and I know for a fact that my edits are accurate, as I was there when these...
4793
kerravon86
Jul 27, 2010 2:06 am
... Hmmm. I didn't believe that, but it seems you're right about that. I didn't bother testing whether 3-byte or 5-byte hex constants caused alignment on those...
4794
kerravon86
Jul 27, 2010 10:26 am
... I don't know, other than I know that when an application is doing a: FRED B FRED I expect the OS to get control after a certain CPU time, preferably...
4795
kerravon86
Jul 28, 2010 12:48 pm
The bootable disk now has two executables - pload and pdos - and pload is now loading the pdos executable into memory, in order to execute it. pdos is just a...
4796
hercules-os380@yahoog...
Jul 29, 2010 11:32 am
Hello, This email message is a notification to let you know that a file has been uploaded to the Files area of the hercules-os380 group. File :...
4797
kerravon86
Jul 29, 2010 6:37 pm
How does MVS 3.8j get the SVC number? I thought there would be fixed addresses for the interrupts, but there doesn't seem to be, suggesting that the number...
4798
Josef 'Jeff' ...
sipek_jeff
Jul 29, 2010 6:46 pm
The architecture defines a byte (if I remember correctly) that's in the SVC instruction itself. The value is stored somewhere in the PSA. Of course, one could...
4799
Tony Harminc
tzha1
Jul 29, 2010 6:53 pm
... Are you asking how to determine which SVC number was issued after the interrupt has taken place? That's not MVS-specific - it's part of the machine...
4800
Gerhard Postpischil
gerhard_post...
Jul 29, 2010 7:29 pm
... Take the SVC Old PSW - if it's in BC mode, the low three bytes have the 24 bit address of the next instruction (i.e., the number is one byte prior). If EC...
4801
kerravon86
Jul 29, 2010 7:55 pm
... It wouldn't have worked, because MVS 3.8j didn't have knowledge of AM 31. However, that was based on an incorect assumption about how things worked. Now I...
4802
kerravon86
Jul 29, 2010 8:47 pm
... Looks like I forgot to set 31-bit mode on, on the interrupt address. (GOTSVC). Which would explain why things only screwed up when returning to C. Will try...
4803
PeterH
peterh5322
Jul 29, 2010 10:44 pm
... It is an "immediate" field, just like the "immediate" field of an MVI instruction, except the instruction length is 2 bytes (same as an RR instruction). ...
4804
Josef 'Jeff' ...
sipek_jeff
Jul 29, 2010 10:53 pm
... Right. I was trying to remember if "immediate" was the name for in on the mainframe (different architectures call same ideas differently...grumble). I...
4805
PeterH
peterh5322
Jul 29, 2010 11:14 pm
... The distinction is significant where an actual implementation is concerned. Most large mainframes suck 64 bits of instructions and data into the I-Unit...
4806
Tony Harminc
tzha1
Jul 30, 2010 12:09 am
... I thought IBM also used those terms. More recently they have an R-unit... ... I'm really not understanding this bit. What has MVS's SVC Screening to do...
4807
PeterH
peterh5322
Jul 30, 2010 12:19 am
... In the 1970s, MVS was taking a 'rupt ... on average ... every 70 microseconds....
4808
kerravon86
Jul 30, 2010 3:03 pm
... A lot of progress made in the just-uploaded version. It now takes a standard MVS executable and noisily processes the SVCs. Obviously mostly stubbed. I ...
4809
kerravon86
Jul 30, 2010 11:53 pm
Ok, I've uploaded a new version that is at least acknowledging the "hello world" from a standard MVS executable. See the "welcome to pcomm" below (equivalent...
4810
kerravon86
Jul 31, 2010 12:18 am
... I'm also happy with the 1-second IPL time. At that point, I'm expecting the sysprog/operator to be able to do a "dir" etc so that they can navigate to the...
4811
Tony Harminc
tzha1
Jul 31, 2010 12:25 am
... SVC 3. ... What are you currently doing upon termination? One of the things often not well understood by people starting out with MVS and its precursors is...
4812
kerravon86
Jul 31, 2010 2:02 am
... Thanks. ... I unwind everything all the way back to the stand-alone loader, where a wait state is entered. The loader isn't really expecting anything to...
4813
Gerhard Postpischil
gerhard_post...
Jul 31, 2010 6:21 am
... Once you leave supervisor state, the only way to get back is via an interrupt. MVS uses SVC 107 (MODESET macro) to allow authorized programs to switch...
4814
kerravon86
Jul 31, 2010 11:53 pm
... Ok. ... Ok, that stuff is easy enough - I'm using MVS standard. ... I personally prefer to debug using printfs rather than looking through any type of...
