EFIDynoTuning Forum

jsa

The address / inline scalar issue...
http://www.efidynotuning.com/forum/view ... 678#p14506

jsa

By coincidence (or intentional obfuscation lol) scalar value happens to be 2280 at address 943a9 and function address happens to be bank 1 2280.

Both are valid with nothing to do with each other than the number 2280.

The example above gives me the scalar name and function name where expected. Post example code it you get something different.

I have put forward the idea, to tvrfan, of SYN as in single symbol.
To generate a definition from dir, the address is needed.
That might look like;
SYN 943A9 "1RATCH_MIN"
SYN 943AF "2RATCH_MIN"
SYM 12280 "Fn0x12280......"
FUN 12280 .......
For now use range for the scalar and function as needed.

What scalars and what files are you referring too?

jsa

Thank you for your continued effort.
4.07.16b 11 Apr 2022 includes support word size tables.

Yes, 4.012 sym addresses can have issues with inline scalars.

jsa

Originally 2280 was for V4.07.16B and 12280 was for 4.012.
The repeats can be deleted.

Put a bank prefix in, though it is technically incorrect it will name the scalars with V4.012.

Code: Select all

SYM  12280 943A8 943B1 "RATCH_MIN"     #W via SADx

jsa

BOOSTEDEVERYTHING wrote: ↑2024 Apr 09, 16:03 I do have another question that I can not figure out. In RZASAMY.dir I have below...
Code: Select all
#SYM  12280 "RATCH_MIN"                 #W via SADx
#SYM 12280 "RATCH_MIN"                 #W via SADx
SYM  144C0 "RATIV"
#SYM 144C0 "RATIV"                     #UW Via SADx

Ah yes, sorry, it is commented out because it's not giving desired results. Out of order too. Something I've been testing with. It needs to have a range applied before un-commenting.

That puts RATCH_MIN

Code: Select all

9439b: a0,37,36           ldw   R36,R136         TMP2L = RATCH;
9439e: a0,23,38           ldw   R38,R122         TMP3L = TP;
943a1: a1,00,3c,3a        ldw   R3a,3c00         TMP4L = 3c00;
943a5: ef,04,e9           call  92cac            Sub_92cac ();
943a8: 89,80,22,42        cmpw  R42,2280         
943ac: db,06              jc    943b4            if (TMP8L < 2280)  {
943ae: a1,80,22,42        ldw   R42,2280         TMP8L = Func_12280;
943b2: 20,0a              sjmp  943be            goto 943be; }

Opcodes a1 and 89 are immediate address modes, they could be addresses or inline scalars, in this case they are inline scalars. SAD is unable to distinguish where to apply SYM's when a scalar value coincides with an 2address. The catch 22 previously discussed. It is something on the fix wish list with @TVRfan
For the purpose of converting DIR to DEF, value 2280 at both addresses 943a9 & 943af need to be captured.

Try

Code: Select all

SYM   2280 943a8 943b1 "RATCH_MIN"                 #W via SADx

Put in the correct address order.
Depending on which version of SAD is used, SAD may complain about a bank number.

You can apply range/s to the function so that it does not name the scalar value 2280 at L943ae.

jsa

Yes, :B replaces :T+.

Current SAD versions are known by version number, but the date is also printed in the LST header for current SAD versions.
These two are the most useable versions at the moment.

Code: Select all

SAD Version 4.0.7.16b (11 Apr 2022)
SAD Version 4.0.12a (12 Mar 2023)

Version 5.08 is the latest but is not quite ready for full time use.

https://github.com/tvrfan/EEC-IV-disass ... ment/Win32
https://github.com/jsa2340/EEC-IV-disas ... evelopment

jsa

You have via SADx as comments, what was the process that involved SADX?

jsa

Sounds like a combination of lookups, named ranges and drop lists would cover it. The build type print out would need formulas to display the required output for the selection choice.

No I'm not inclined to get involved with it.

jsa

Thanks.

It's an ugly spreadsheet that does the jump comments, sometimes it works ok'ish other times it fights all the way. For omae2 it wanted to do +E0 scientific notation math for any address with e followed by numbers in it instead of text. Ended up doing a pile of manual fixes.

Comparison spreadsheets can be handy for guiding SAD development.

jsa

Branch comment file;

Stock_EKO2_256k_CMT.zip: (117.16 KiB) Downloaded 10 times

EFIDynoTuning Forum

Search found 271 matches

Re: SAD Version 5 (alpha)

Re: EEC V file conversion

Re: SAD Version 5 (alpha)

Re: EEC V file conversion

Re: EEC V file conversion

Re: SAD Directive Formatting

Re: EEC V file conversion

Re: EEC V file conversion

Re: EEC V file conversion

Re: EEC V file conversion