gmane.comp.emulators.hercules390.general

Re: Basic bootstrap loader program disassembled

BruceTSmith — 2008-12-05T03:30:36

The CCW "Assembler Instruction" has been included in EVERY Assembler, for EVERY S/3x0 OS, since, well since day 1. :) Here's the oldest Asm reference I can find online, "IBM System/360 BPS BAL", Feb 1965. CCW is described on Page 30 (PDF frame 29)... http://www.bitsavers.org/pdf/ibm/360/C20-6503-0_BAL_Feb65.pdf BTS :)))

Re: Basic bootstrap loader program disassembled

rhtatum — 2008-12-04T23:26:27

Lindy, It's not a macro at all;if you go to GC33-4010, File No. S370-21 (OS/VS, DOS/VS), the OS/VS and DOS/VS) Assembler Language manual which applied to VS1 Release 2, VS2 Release 1 and DOS/VS Release 28, Section G: Symbol and Data Definition (or the old Assembler F Programmer's Guide and a plethora of other places!!), you will find the following: G30 -- the ccw instruction Purpose You can use the CCW instruction to define and generate an eight-byte channel command word for input/output operations. The channel command word is eight-byte field aligned at a doubleword boundary, and contains the information described in the figure to the right. [I won't even try to draw a picture, but the thingies are as follows] Bits 0-7: the command code bits 8-31: Address of data to operate upon bits 32-36: Flags Bits 37-39: Must be specified as Zeros Bits 40-47: Set to Zeros by Assembler Bits 48-63: Byte Count or Length of Data Specifications The format of the CCW instruction statement is given in the figure to the right

RE: Basic bootstrap loader program disassembled

Lindy Mayfield — 2008-12-04T22:35:00

Ah yes, they are assembler instructions. I found them in the hlasm reference. That's good to know, thanks. From: hercules-390-hHKSG33TihhbjbujkaE4pw< at >public.gmane.org [mailto:hercules-390-hHKSG33TihhbjbujkaE4pw< at >public.gmane.org] On Behalf Of Josef 'Jeff' Sipek Sent: 5. joulukuuta 2008 0:28 To: hercules-390-hHKSG33TihhbjbujkaE4pw< at >public.gmane.org Subject: Re: [hercules-390] Basic bootstrap loader program disassembled A quick web search releaved: Developing z/OS applications and Web service interfaces (not sure about the pub #), which has the following: CCW and CCW0 Instructions The CCW and CCW0 instructions define and generate an 8-byte, format-0 channel command word for input/output operations. A format-0 channel command word allows a 24-bit data address. The CCW and CCW0 instructions have identical functions. If a control section has not been established, CCW and CCW0 will initiate an unnamed (private) control section. '-symbol-' '-CCW0-' ... The following are examples of CCW and CCW0 statements:

Re: Basic bootstrap loader program disassembled

Josef 'Jeff' Sipek — 2008-12-04T22:27:52

A quick web search releaved: Developing z/OS applications and Web service interfaces (not sure about the pub #), which has the following: CCW and CCW0 Instructions The CCW and CCW0 instructions define and generate an 8-byte, format-0 channel command word for input/output operations. A format-0 channel command word allows a 24-bit data address. The CCW and CCW0 instructions have identical functions. If a control section has not been established, CCW and CCW0 will initiate an unnamed (private) control section. '-symbol-' '-CCW0-' ... The following are examples of CCW and CCW0 statements: WRITE1 CCW 1,DATADR,X'48',X'50' WRITE2 CCW0 1,DATADR,X'48',X'50' Jeff.

RE: Basic bootstrap loader program disassembled

Lindy Mayfield — 2008-12-04T22:19:32

Maybe you mean IOSDCCW in SYS1.MODGEN? There is also a DEFCCW in SYS1.MACLIB. From: hercules-390-hHKSG33TihhbjbujkaE4pw< at >public.gmane.org [mailto:hercules-390-hHKSG33TihhbjbujkaE4pw< at >public.gmane.org] On Behalf Of Josef 'Jeff' Sipek Sent: 4. joulukuuta 2008 23:36 To: hercules-390-hHKSG33TihhbjbujkaE4pw< at >public.gmane.org Subject: Re: [hercules-390] Basic bootstrap loader program disassembled For what its worth, there's a CCW macro that lets you construct channel programs more tersely. I've never used it (never had to) but just pointing out that it's there. Jeff. -- Only two things are infinite, the universe and human stupidity, and I'm not sure about the former. - Albert Einstein [Non-text portions of this message have been removed]

Re: Basic bootstrap loader program disassembled

Josef 'Jeff' Sipek — 2008-12-04T21:35:51

For what its worth, there's a CCW macro that lets you construct channel programs more tersely. I've never used it (never had to) but just pointing out that it's there. Jeff.

Basic bootstrap loader program disassembled

Lindy Mayfield — 2008-12-04T21:31:04

Probably people remember helping me understand how the IPL / boot process works on s/3x0 systems. A few suggested that I disassemble and comment a basic bootstrap loader program found on VM called 3card loader s1. Below is a first cut at it, including a few suggestions from people on the assembler list. http://www.lilliana.eu/downloads/3card.loader.txt It is actually a very nice piece of code. So this gets me halfway through my "hello world" challenge. It might seem like the rest is easy, just write a couple of lines of assembler, punch it after 3card loader and run it. Well, easy for some maybe when you know it. (-: I'm still working on that part. Thanks for everyone's help with this. Lindy [Non-text portions of this message have been removed]

Re: Three big problems running hercules

kulovits_christian — 2008-12-04T10:19:13

Hello Roger, my fishGui is 1.11.1.5265, which is the last available built. regards, Christian

Re: AW: AW: HHCTU002E Error opening TUN/TAP device: 00-14-22-BC-54-A0

2008-12-04T10:12:46

Hi Peter, I did try to run the tt32test in DOS prompt with "run as administrator" selected, but still get the output as below; 13:33:10.531 ** CFishPackApp::FishPackGetWinPCapInfo: CreateProcess(FWPCUtil) failed; rc=14001 (0x000036B1): The application has failed to start because its side-by-side configuration is incorrect. Please see the application event log for more detail. 13:33:10.532 Begin TT32Test.exe, version: "3.2.1.160" (3.2.1.160): "Release version" ... 13:33:10.532 Copyright (C) 2002-2007, Software Development Laboratories (aka 'Fish' (David B. Trout)). 13:33:10.532 13:33:10.532 Running on Windows 2000, version 6.0.6001 Service Pack 1 13:33:10.532 13:33:10.532 ERROR: ************************************************************** 13:33:10.532 ERROR: ** ** 13:33:10.532 ERROR: ** *** ERROR *** ** 13:33:10.532 ERROR: ** ** 13:33:10.532 ERROR:

Re: Re: CDROM emulation bug

Harold Grovesteen — 2008-12-04T08:00:16

Thanks Paul for validating my interpretation of what I was seeing. I was unsure exactly how to fix it, though. Thanks Jan for fixing it. The beauty of many eyes and open source! Harold Grovesteen pa3efu wrote:

Re: CDROM emulation bug

pa3efu — 2008-12-04T06:21:14

Argh... You are right! I wrote this 8 years ago, did not see it then, did not see it now. We will fix it Thanks! jj --- In hercules-390-hHKSG33TihhbjbujkaE4pw< at >public.gmane.org, "kerravon86" ...> wrote: we loc

AW: AW: HHCTU002E Error opening TUN/TAP device: 00-14-22-BC-54-A0

Peter Glanzmann — 2008-12-04T05:41:01

Hi Lim In my case tt32test did also not work (it did not found any network adapters). It only worked, when I ran it with administrator priviledges. Maybe you can try this first. If tt32test works correctly with administrative rights, then you can assume also some authorization problems running Hercules. If you have still the same result, then probably it is not an authorization problem. In this case your problem seems not to be the same that I had... Kind regards, Peter -----Ursprüngliche Nachricht----- Von: hercules-390-hHKSG33TihhbjbujkaE4pw< at >public.gmane.org [mailto:hercules-390-hHKSG33TihhbjbujkaE4pw< at >public.gmane.org] Im Auftrag von limmi01-GTSHjikGRxpWk0Htik3J/w< at >public.gmane.org Gesendet: Mittwoch, 3. Dezember 2008 10:50 An: hercules-390-hHKSG33TihhbjbujkaE4pw< at >public.gmane.org Betreff: Re: AW: [hercules-390] HHCTU002E Error opening TUN/TAP device: 00-14-22-BC-54-A0 Hi Dave, Which TUN/TAP program you are referring to ? Pardon my ignorance on this. Are u refering to tt32test.exe ? Please elaborate. Th

Re: Three big problems running hercules

August Treubig — 2008-12-04T02:11:32

Yes, Roger, I was talking about the S0C9 problem. Hopefully IBM will do something with the open APAR. Aug --- In hercules-390-hHKSG33TihhbjbujkaE4pw< at >public.gmane.org, Roger Bowler ...> wrote:

Blank Lines in GUI -> was Re: Re: Three big problems running hercules

Dave Wade — 2008-12-04T00:15:03

Roger Bowler wrote: Roger, I get blank lines in the GUI on my work laptop. I have an up-to-date GUI and a Hercules snapshot from May. I don't get it on my personal laptop. I'll see if I can figure out why. Dave.

Re: Re: Three big problems running hercules

Roger Bowler — 2008-12-04T00:10:06

August, I think you are talking about the 0C9 problem. When I mentioned upgrading to the latest GUI, I was talking about the "blank lines" problem which Christian said only occured when he was using the GUI. There are actually three different problems being discussed here. It might be clearer if each problem had a separate message with a distinct title.

Re: Three big problems running hercules

August Treubig — 2008-12-03T23:53:32

Roger, I have had it happen running on Linux (no gui involved). On about 4 occasions on 2 different machines. Both were running Centos Linux 5.2. zOS is 1.8. The hercules is the CVS version from early August. So, I don't see how the Windows Gui is involved, unless I misinterpreted what you meant. August --- In hercules-390-hHKSG33TihhbjbujkaE4pw< at >public.gmane.org, Roger Bowler ...> wrote: STCK defined. the ARE NO

Re: Re: Three big problems running hercules

Roger Bowler — 2008-12-03T23:42:09

Interesting! This sounds suspiciously like a TOD clock problem. Two STCK returning the same value would give an unexpected zero difference, resulting in an 0C9 if that difference is used as a divisor. This is only a guess from reading the apar description and vaguely remembering similar problems in the past. I wonder if you might have an older version of the gui which does not match the current version of Hercules? Make sure you get the latest version of the GUI from fish's site.

Re: CDROM emulation bug

kerravon86 — 2008-12-03T21:12:25

And you want to know if you had a partial read, so that you can terminate the loop. You've now clobbered your ability to do that partial read check. So if you wanted to keep the existing structure, you'd need something like this: C:\devel\hercules>cvs diff -c ipl.c Index: ipl.c =================================================================== RCS file: c:\cvsroot/hercules/ipl.c,v retrieving revision 1.1.1.1 diff -c -r1.1.1.1 ipl.c *** ipl.c 16 Nov 2007 14:44:26 -0000 1.1.1.1 --- ipl.c 3 Dec 2008 21:02:31 -0000 *************** *** 612,620 **** } pageaddr += PAGEFRAME_PAGESIZE; pageaddr &= PAGEFRAME_PAGEMASK; - pagesize = PAGEFRAME_PAGESIZE; } ! while (len == (int)pagesize); close(fd); --- 612,619 ---- } pageaddr += PAGEFRAME_PAGESIZE; pageaddr &= PAGEFRAME_PAGEMASK; } ! while (len == (int)pagesize && (pagesize = PAGEFRAME_PAGESIZE,1)); close(fd); That's not how I would do it tho

Re: CDROM emulation bug

pa3efu — 2008-12-03T19:27:46

Hmmm, not sure if I see the bug here yet. We first calculate how large the remainder of the first page is (which is potentially loaded): pagesize = PAGEFRAME_PAGESIZE - (startloc & PAGEFRAME_BYTEMASK); We then set the startaddress where we start loading: pageaddr = startloc; We then start a loop: do We check for the storage limit: if (pageaddr >= sysblk.mainsize) logmsg(_("HHCCP034W load_main: terminated at end of mainstor\n")); Then we attempt to read upto pagesize bytes into storage: len = read(fd, sysblk.mainstor + pageaddr, pagesize); If we did read anything then we update reference+ change bits of the storage frame: if (len > 0) STORAGE_KEY(pageaddr, &sysblk) |= STORKEY_REF|STORKEY_CHANGE; We advande to the next page: pageaddr += PAGEFRAME_PAGESIZE; We reset to the page next boundary...: pageaddr &= PAGEFRAME_PAGEMASK; As all pages read will be 4k long (other then the first page as we will have only read the remainder of that page if the starting loc was not page aligned): page

CDROM emulation bug

Harold Grovesteen — 2008-12-03T11:53:57

This code from ARCH_DEP(load_main) in ipl.c has some interesting results when the start location for a file in the .ins file is not on a page boundary. I am not a C expert, but if I am reading the code correctly, the while condition fails when the start location is not on a page boundary and the loop stops after the first read of the initial partial page. For a start location not on a page boundary the len resulting from the read is the size of the partial block, however, when the while condition is checked, the pagesize value has been updated to that of a full page. Now len and pagesize do not match and the load stops. I found this while testing with Sirius. When I saved and loaded the image file on a page boundary, all got better. I have a work around, but this looks, feels and smells like a bug. pagesize = PAGEFRAME_PAGESIZE - (startloc & PAGEFRAME_BYTEMASK); pageaddr = startloc; do { if (pageaddr >= sysblk.mainsize) { logmsg(_("HHCCP034W load_main:

Re: AW: HHCTU002E Error opening TUN/TAP device: 00-14-22-BC-54-A0

2008-12-03T09:49:52

Hi Dave, Which TUN/TAP program you are referring to ? Pardon my ignorance on this. Are u refering to tt32test.exe ? Please elaborate. Thank you. Regards Lim ML ----- Original Message ----- From: Dave Wade public.gmane.org> Date: Wednesday, December 3, 2008 3:58 pm Subject: Re: AW: [hercules-390] HHCTU002E Error opening TUN/TAP device: 00-14-22-BC-54-A0 To: hercules-390-hHKSG33TihhbjbujkaE4pw< at >public.gmane.org [Non-text portions of this message have been removed]