|
Hi ? I¡¯m using Hyperion 4.5 as below HHC01413I Hercules version 4.5.0.10830-SDL-g58578601????????????????????????????????????? ?????????????????????????????? HHC01414I (C) Copyright 1999-2022 by Roger Bowler, Jan Jaeger, and others?????????????????????????????????????????????? HHC01417I ** The SoftDevLabs version of Hercules **?????????????????????????????????????????????????????? ?????????????? HHC01415I Build date: Nov 26 2022 at 16:25:41?????????????????????????????????????????????????????????????????????????? HHC01417I Built with: Microsoft Visual Studio 2008 (MSVC 150030729 1)?????????????????????????????????????????????????? HHC01417I Build type: Windows MSVC AMD64 host architecture build??????????????????????????????????????????????????????? ? All my dasd devices are defined as 3390 with CU=3990-6. Everything works fine but this message below is always present for all my devices after IPL z/os. ? HHC01315I 0:2107 CHAN: ccw F760000C 7E84DF58=>18000000 00004100 00000000????????? ............????????????????????????? HHC01312I 0:2107 CHAN: stat 0E00, count 0000??????????????????????????????????????????????????????????????????????????? HHC01313I 0:2107 CHAN: sense 80000000 00FFFF01 00000000 00000000 00000000 00000000 00000080 0002BF0B??????????????????? HHC01314I 0:2107 CHAN: sense CMDREJ??????????? ? Even using old Hyperion versions the same always occurred. What is wrong and what can be done to avoid this situation? ? TIA Jo?o ? ? -- att. Jo?o
|
It's not unusual for the system to issue channel commands in the expectation they will fail. It does this to see if particular capabilities are present. You can ignore the error if it does not cause problems for the running system.
Hi ? I¡¯m using Hyperion 4.5 as below HHC01413I Hercules version 4.5.0.10830-SDL-g58578601????????????????????????????????????? ?????????????????????????????? HHC01414I (C) Copyright 1999-2022 by Roger Bowler, Jan Jaeger, and others?????????????????????????????????????????????? HHC01417I ** The SoftDevLabs version of Hercules **?????????????????????????????????????????????????????? ?????????????? HHC01415I Build date: Nov 26 2022 at 16:25:41?????????????????????????????????????????????????????????????????????????? HHC01417I Built with: Microsoft Visual Studio 2008 (MSVC 150030729 1)?????????????????????????????????????????????????? HHC01417I Build type: Windows MSVC AMD64 host architecture build??????????????????????????????????????????????????????? ? All my dasd devices are defined as 3390 with CU=3990-6. Everything works fine but this message below is always present for all my devices after IPL z/os. ? HHC01315I 0:2107 CHAN: ccw F760000C 7E84DF58=>18000000 00004100 00000000????????? ............????????????????????????? HHC01312I 0:2107 CHAN: stat 0E00, count 0000??????????????????????????????????????????????????????????????????????????? HHC01313I 0:2107 CHAN: sense 80000000 00FFFF01 00000000 00000000 00000000 00000000 00000080 0002BF0B??????????????????? HHC01314I 0:2107 CHAN: sense CMDREJ??????????? ? Even using old Hyperion versions the same always occurred. What is wrong and what can be done to avoid this situation? ? TIA Jo?o ? ? -- att. Jo?o
|
Jo?o Reginato wrote: [...] Everything works fine but this message below is always present
for all my devices after IPL z/os.
HHC01315I 0:2107 CHAN: ccw F760000C 7E84DF58=>18000000 00004100 00000000
HHC01312I 0:2107 CHAN: stat 0E00, count 0000
HHC01313I 0:2107 CHAN: sense 80000000 00FFFF01 00000000 00000000 00000000
00000000 00000080 0002BF0B
HHC01314I 0:2107 CHAN: sense CMDREJ [...] What is wrong and what can be done to avoid this situation? What is wrong? Nothing. What can be done to avoid it? Nothing! :) Nothing is really wrong with Hercules OTHER than possibly the fact that it does not currently support that particular CCW opcode. If we could locate documentation somewhere that explains what that particular CCW opcode is (what it does and how it's used), then we *might* be able to MAYBE add support for it. (Maybe) But since we don't have the documentation for it and thus have no idea what it does or how it's used, that makes it kind of difficult/impossible for us to support it. ;-) As Jay Maynard explained in his reply, it's not uncommon for IBM to update their operating system to issue new commands to devices that THEY (IBM) know are supported (by *real* 3390s) but which are not publicly documented anywhere. THEY (IBM) probably have internal documentation that explains it, but it's not something they wish to make public for whatever reason. As you've already noticed however, supporting it is apparently not that critical though, since the system appears to run just fine without it, so my only advice to you would be to simply ignore it. -- "Fish" (David B. Trout) Software Development Laboratories mail: fish@...
|
Ok thank you -----Mensagem original----- De: [email protected] < [email protected]> Em nome de Fish Fish Enviada em: sexta-feira, 6 de janeiro de 2023 17:18 Para: [email protected]Assunto: Re: [H390-MVS] 3390/3990-06 I/O Error reported Jo?o Reginato wrote: [...] Everything works fine but this message below is always present for all
my devices after IPL z/os.
HHC01315I 0:2107 CHAN: ccw F760000C 7E84DF58=>18000000 00004100
00000000 HHC01312I 0:2107 CHAN: stat 0E00, count 0000 HHC01313I 0:2107
CHAN: sense 80000000 00FFFF01 00000000 00000000 00000000
00000000 00000080 0002BF0B
HHC01314I 0:2107 CHAN: sense CMDREJ [...] What is wrong and what can be done to avoid this situation? What is wrong? Nothing. What can be done to avoid it? Nothing! :) Nothing is really wrong with Hercules OTHER than possibly the fact that it does not currently support that particular CCW opcode. If we could locate documentation somewhere that explains what that particular CCW opcode is (what it does and how it's used), then we *might* be able to MAYBE add support for it. (Maybe) But since we don't have the documentation for it and thus have no idea what it does or how it's used, that makes it kind of difficult/impossible for us to support it. ;-) As Jay Maynard explained in his reply, it's not uncommon for IBM to update their operating system to issue new commands to devices that THEY (IBM) know are supported (by *real* 3390s) but which are not publicly documented anywhere. THEY (IBM) probably have internal documentation that explains it, but it's not something they wish to make public for whatever reason. As you've already noticed however, supporting it is apparently not that critical though, since the system appears to run just fine without it, so my only advice to you would be to simply ignore it. -- "Fish" (David B. Trout) Software Development Laboratories mail: fish@... -- att. Jo?o
|
Tommy Sprinkle may be able to help...
Jo?o Reginato wrote:
[...]
> Everything works fine but this message below is always present
> for all my devices after IPL z/os.
>
> HHC01315I 0:2107 CHAN: ccw F760000C 7E84DF58=>18000000 00004100 00000000
> HHC01312I 0:2107 CHAN: stat 0E00, count 0000
> HHC01313I 0:2107 CHAN: sense 80000000 00FFFF01 00000000 00000000 00000000
> 00000000 00000080 0002BF0B
> HHC01314I 0:2107 CHAN: sense CMDREJ
[...]
> What is wrong and what can be done to avoid this situation?
What is wrong? Nothing.
What can be done to avoid it? Nothing! :)
Nothing is really wrong with Hercules OTHER than possibly the fact that it does not currently support that particular CCW opcode.
If we could locate documentation somewhere that explains what that particular CCW opcode is (what it does and how it's used), then we *might* be able to MAYBE add support for it. (Maybe)
But since we don't have the documentation for it and thus have no idea what it does or how it's used, that makes it kind of difficult/impossible for us to support it. ;-)
As Jay Maynard explained in his reply, it's not uncommon for IBM to update their operating system to issue new commands to devices that THEY (IBM) know are supported (by *real* 3390s) but which are not publicly documented anywhere. THEY (IBM) probably have internal documentation that explains it, but it's not something they wish to make public for whatever reason.
As you've already noticed however, supporting it is apparently not that critical though, since the system appears to run just fine without it, so my only advice to you would be to simply ignore it.
--
"Fish" (David B. Trout)
Software Development Laboratories
mail: fish@...
|
Theodore M Rolle Jr wrote: Tommy Sprinkle may be able to help... How? -- "Fish" (David B. Trout) Software Development Laboratories mail: fish@...
|
No offense, but I think Tommy has enough on his plate...
Joe
On Fri, Jan 6, 2023 at 4:57 PM Theodore M Rolle Jr < stercor@...> wrote: Tommy Sprinkle may be able to help...
Jo?o Reginato wrote:
[...]
> Everything works fine but this message below is always present
> for all my devices after IPL z/os.
>
> HHC01315I 0:2107 CHAN: ccw F760000C 7E84DF58=>18000000 00004100 00000000
> HHC01312I 0:2107 CHAN: stat 0E00, count 0000
> HHC01313I 0:2107 CHAN: sense 80000000 00FFFF01 00000000 00000000 00000000
> 00000000 00000080 0002BF0B
> HHC01314I 0:2107 CHAN: sense CMDREJ
[...]
> What is wrong and what can be done to avoid this situation?
What is wrong? Nothing.
What can be done to avoid it? Nothing! :)
Nothing is really wrong with Hercules OTHER than possibly the fact that it does not currently support that particular CCW opcode.
If we could locate documentation somewhere that explains what that particular CCW opcode is (what it does and how it's used), then we *might* be able to MAYBE add support for it. (Maybe)
But since we don't have the documentation for it and thus have no idea what it does or how it's used, that makes it kind of difficult/impossible for us to support it. ;-)
As Jay Maynard explained in his reply, it's not uncommon for IBM to update their operating system to issue new commands to devices that THEY (IBM) know are supported (by *real* 3390s) but which are not publicly documented anywhere. THEY (IBM) probably have internal documentation that explains it, but it's not something they wish to make public for whatever reason.
As you've already noticed however, supporting it is apparently not that critical though, since the system appears to run just fine without it, so my only advice to you would be to simply ignore it.
--
"Fish" (David B. Trout)
Software Development Laboratories
mail: fish@...
|
He writes CCW programs in Assembler language. Documented, too! He uses MVS 3.8 but it wouldn't hurt to ask....
Theodore M Rolle Jr wrote:
> Tommy Sprinkle may be able to help...
How?
--
"Fish" (David B. Trout)
Software Development Laboratories
mail: fish@...
|
He's also part of the LzLabs group thats being sued by IBM over patent/copyright infringement.
Joe
On Fri, Jan 6, 2023 at 6:18 PM Theodore M Rolle Jr < stercor@...> wrote: He writes CCW programs in Assembler language. Documented, too! He uses MVS 3.8 but it wouldn't hurt to ask....
Theodore M Rolle Jr wrote:
> Tommy Sprinkle may be able to help...
How?
--
"Fish" (David B. Trout)
Software Development Laboratories
mail: fish@...
|
Joe Monk wrote: He's also part of the LzLabs group thats being sued by IBM
over patent/copyright infringement. It's unclear (to me) whether he actually works FOR LzLabs, or AT LzLabs (i.e. whether he is directly employed by them, or is simply doing contract work for them). His LinkedIn profile says he's self employed. -- "Fish" (David B. Trout) Software Development Laboratories mail: fish@...
|
People ? If you look at the POP, the chapter on ¡®basic I/O ?functions¡¯, there is a chart (in my POP, fig 15-5) that shows the bit assignment for CCW command codes.? The particular code in question is X¡¯18.¡¯ If this is a fmt 0 CCW, that is supposed to be interpreted as a TIC (which is valid) and in the case of a fmt 1 CCW, it is invalid. The message does not indicate the CCW format being used (it should) but, if it is a fmt 1 CCW, the message is correct. IBM (and anybody else) cannot add arbitrary CCW command codes here ¨C the figure indicates where they are allowed. So, this is not a ¡®new¡¯ command code that Hercules cannot understand, it is an error!?? ? Peter Morrison
|
The CCW in question:?18000000 00004100 00000000
Thats not a valid format-0 CCW either.
The address if it were a format-0 CCW is absolute location zero.
Either way its not a valid TIC CCW.
Joe
On Sat, Jan 7, 2023 at 4:29 PM < petmo@...> wrote: People ? If you look at the POP, the chapter on ¡®basic I/O ?functions¡¯, there is a chart (in my POP, fig 15-5) that shows the bit assignment for CCW command codes.? The particular code in question is X¡¯18.¡¯ If this is a fmt 0 CCW, that is supposed to be interpreted as a TIC (which is valid) and in the case of a fmt 1 CCW, it is invalid. The message does not indicate the CCW format being used (it should) but, if it is a fmt 1 CCW, the message is correct. IBM (and anybody else) cannot add arbitrary CCW command codes here ¨C the figure indicates where they are allowed. So, this is not a ¡®new¡¯ command code that Hercules cannot understand, it is an error!?? ? Peter Morrison
|
Peter Morrison wrote: [...] The particular code in question is X¡¯18.¡¯ Wrong. The particular code in question is X'F7'. Refer to Jo?o's original post and my reply: * /g/H390-MVS/message/3187 * /g/H390-MVS/message/3192The CCW that is failing is F760000C 7E84DF58. The DATA that the ccw points to is 18000000 00004100 00000000. [...] So, this is not a ¡®new¡¯ command code Whether it is a new or old command code makes no difference. The only thing that matters is that it is an UNDOCUMENTED command code. that Hercules cannot understand, it is an error! It is a LIMITATION, not an error (per se). -- "Fish" (David B. Trout) Software Development Laboratories mail: fish@...
|
Joe Monk wrote: The CCW in question: 18000000 00004100 00000000 Wrong. The CCW in question is F760000C 7E84DF58. -- "Fish" (David B. Trout) Software Development Laboratories mail: fish@...
|
For the records: On z/OS 2.5, the "limitation" appears for every dasd during the IPL, with no notable effect, but also for a init volume ICKDSF, that abends with U0028:
HHC01315I 0:0303 CHAN: ccw F760000C 7E099F58=>18000000 00004100 00000000? ? ? ? ? ............
HHC01312I 0:0303 CHAN: stat 0E00, count 0000
HHC01313I 0:0303 CHAN: sense 80000000 00FFFF01 00000000 00000000 00000000 00000000 00000080 00000000
HHC01314I 0:0303 CHAN: sense CMDREJ? ?
|
After a bit of research, I found a small clue: At??can see the F7 CCW cmd as:
#define DASD_ECKD_CCW_DSO 0xF7
At??that cmd is used:
ccw->cmd_code = DASD_ECKD_CCW_DSO;
But, it is for sub-order 0x81, not 0x18.
ras_data->order = DSO_ORDER_RAS;
The GTF trace in the z/OS, shows a chained F7+3E (SDO + Read subsystem data):
CCW CHAIN? ? ? ? ?FORMAT 1? ? ? ? ? SSCH? ? ? ? ? ? ?
023A5D80 F760000C 7E0ECF58 18000000 00004100 00000000
023A5D88 3E200100 7E0ECDF8? ? ? ? ? ? ? ? ? ? ? ? ? ?
I think that the linux/s390 source code can be a great ally for Hercules, but in this case it does not let us see further.
P.S. Hercules ckddasd.c uses simple hex constants:
case 0x3E:?
Should Hercules developers copy those beautiful @define ECKD constants from linux?
|
Jo?o Reginato
Theodore M Rolle Jr