Re: Analyzing SAV save files from a PC

If you use SST to dump the pgm that displays save files, QSYS/QSRDSAVF,
you see some interesting tables (from a V5R1 machine, folded to 16 chars
for email):

0A000A02E2C1E5D3  C9C24040404003E2  *....SAVLIB    .S*
C1E5E2E8E2404040  4004E2C1E5E2E8E2  *AVSYS    .SAVSYS*
4040404005E2C1E5  E2E8E24040404006  *    .SAVSYS    .*
E2C1E5E2E8E24040  404007E2C1E5C3C6  *SAVSYS    .SAVCF*
C740404040080AE2  C1E5E2C5C3C4E3C1  *G    ..SAVSECDTA*
400BE2C1E5C4D3D6  404040400CE2C1E5  * .SAVDLO    .SAV*
E2C1E5C6C4E3C10D  E2C1E5D3C9C3D7C7  *SAVFDTA.SAVLICPG*
D4400E0FE2C1E5E2  E3C74040404010E2  *M ..SAVSTG    .S*
C1E5D6C2D1404040  4011E2C1E5C3C8C7  *AVOBJ    .SAVCHG*
D6C2D14012E2C1E5  E2E8E2C9E2C44013  *OBJ .SAVSYSISD .*
14E2C1E540404040  40404015FFB85CD3  *.SAV       ...*L*
C9C2D34040404040  FFB7F0F1F0F00000  *IBL     ..0100..*
000000000000207F  FF175CE2F3F84040  *..........*S38  *
20E5F1D9F1D4F020  E5F1D9F1D4F221E5  *.V1R1M0.V1R1M2.V*
F1D9F2D4F022E5F1  D9F2D4F023E5F1D9  *1R2M0.V1R2M0.V1R*
F3D4F024E5F2D9F1  D4F025E5F2D9F1D4  *3M0.V2R1M0.V2R1M*
F126E5F2D9F2D4F0  27E5F2D9F2D4F028  *1.V2R2M0.V2R2M0.*
E5F2D9F2D4F029E5  F2D9F2D4F030E5F2  *V2R2M0.V2R2M0.V2*
D9F3D4F031E5F3D9  F0D4F532E5F3D9F1  *R3M0.V3R0M5.V3R1*
D4F033E5F3D9F1D4  F034E5F3D9F1D4F0  *M0.V3R1M0.V3R1M0*
35E5F3D9F6D4F036  E5F3D9F2D4F03AE5  *.V3R6M0.V3R2M0.V*
F3D9F7D4F03BE5F4  D9F1D4F03DE5F4D9  *3R7M0.V4R1M0.V4R*
F1D4F43CE5F4D9F2  D4F03EE5F4D9F3D4  *1M4.V4R2M0.V4R3M*
F03FE5F4D9F4D4F0  40E5F4D9F5D4F041  *0.V4R4M0 V4R5M0.*
E5F5D9F1D4F00EEF  D8E2D94BD4C5D4C2  *V5R1M0..QSR.MEMB*


The first table gives the command byte:

  02 SAVLIB
  03 SAVSYS
  04 SAVSYS
  05 SAVSYS
  06 SAVSYS
  07 SAVCFG
  08 SAVSECDTA
  0A SAVSECDTA
  0B SAVDLO
  0C SAVSAVFDTA
  0D SAVLICPGM
  0E SAVSTG
  0F SAVSTG
  10 SAVOBJ
  11 SAVCHGOBJ
  12 SAVSYSISD
  13 SAV
  14 SAV

Strangely, it lists SAVOBJ as x10, although you get x01 if you do SAVOBJ
to a save file.  If you change the save file to x10 with SST, DSPSAVF
shows "QSYS" for the save command.

The second table shows the release codes:

  20 V1R1M0
  20 V1R1M2
  21 V1R2M0
  22 V1R2M0
  23 V1R3M0
  24 V2R1M0
  25 V2R1M1
  26 V2R2M0
  27 V2R2M0
  28 V2R2M0
  29 V2R2M0
  30 V2R3M0
  31 V3R0M5
  32 V3R1M0
  33 V3R1M0
  34 V3R1M0
  35 V3R6M0
  36 V3R2M0
  3A V3R7M0
  3B V4R1M0
  3C V4R2M0
  3D V4R1M4
  3E V4R3M0
  3F V4R4M0
  40 V4R5M0
  41 V5R1M0

Many anomalies and duplications.  As a programmer, it's comforting to
see that even Rochester doesn't have its act together entirely.

(They seem to have considered it a decimal number until V3R7.)

If you use SST to change the byte in a save file to less than x20,
DSPSAVF gives an error.  Presumably those are S/38 codes.

Anybody have a running S/38? :-)

--Dave

On Wed, 2004-08-25 at 17:58, Mark S. Waterbury wrote: 
> Cool!  IFS was not introduced until V3R1, so you are okay with regard to
> V2R3 and V3R0M5. (V3R0M5 was essentially just V2R3 "patched" to run on the
> then-new black box models. Why didn't they just call it V2R3M1? Because, if
> you recall, IBM charged a significant fee to upgrade from V2 to V3.) If
> anyone is really interested, I have an old machine running V2R1M0, I could
> "fire it up" and create a save file and see what it has for codes for V2R1,
> and I could probably TGTRLS(*PRV) back at least as far as V1R3, perhaps even
> V1R2... ;-)
> 
> ----- Original Message -----
> > From: "Dave McKenzie" <davemck@xxxxxxxxxxxxx>
> > To: "MI Programming on the AS400 / iSeries" <mi400@xxxxxxxxxxxx>
> > Sent: Wednesday, August 25, 2004 5:13 PM
> > Subject: Re: [MI400] Analyzing SAV save files from a PC
> >
> 
> > Here's what I know about the release level codes:
> >
> >          SAVOBJ/
> >          SAVLIB   SAV
> > ---------------------
> > V5R3M0    x43     xAF
> > V5R2M0     42      AE
> > V5R1M0     41      AD
> > V4R5M0     40      AC
> > V4R4M0     3F      AB
> > V4R3M0     3E      AA *
> > V4R1M4     3D      A9 *
> > V4R2M0     3C      A8 *
> > V4R1M0     3B      A7 *
> > V3R7M0     3A      A6 *
> > V3R2M0     36      A2
> > V3R6M0     35      A1 *
> > V3R1M0     34      A0
> > V3R0M5     31
> > V2R3M0     30
> >
> > Notice the gaps and codes out of order.  They may give a glimpse into
> > the order that Rochester worked on the releases.
> >
> > The SAV codes marked with * are guesses, based on the sequence of the
> > SAVOBJ/SAVLIB codes, which I'm sure of.
> >
> > I don't recall when the IFS was introduced, so I didn't try to guess
> > V3R0M5 and V2R3.
> >
> > Don't ask me what V4R1M4 was :-)  I found it by manually changing the
> > byte in the save file with SST, then looking at what DSPSAVF showed.
> >
> > --Dave
> >
> > On Wed, 2004-08-25 at 13:46, James H H Lampert wrote:
> > > As of now, I'm able to generate a list of complete pathnames of
> everything
> > > in the save file. Thanks, Dave.
> > >
> > > Would anybody happen to know what the SAV release level codes are for
> V4R3
> > > and prior? And what about the SAVOBJ/SAVLIB release level codes for
> V5Rx?
> > >
> > > --
> > > JHHL
> >