Subfunctions hyperlinking: NetWare

[ecm-pushbx]

Some of the hyperlinks do use SF= like a register:

TheList/source/Interrupt List/INT 21 DOS Function Calls/INT 21F2 Novell NetWare v301 shell interface MULTIPLEXOR.txt

Line 421 in f35cbd5

7Bh/3Dh Get Server Set Categories (see AX=F27Bh/SF=3Dh)

I found that the NetWare entries use AX = 1234h subfn 56h to list such an entry:

TheList/source/Interrupt List/INT 21 DOS Function Calls/INT 21F27BSF3D Novell NetWare v4 GET SERVER SET CATEGORIES.txt

Line 8 in f35cbd5

AX = F27Bh subfn 3Dh

Open question is whether this pattern holds for all entries referenced using SF= like a register.

(Finding the URLs for those files is unnecessarily complicated due to #49 )

[ecm-pushbx]

Supported by IntList with this changeset: https://hg.pushbx.org/ecm/tractest/rev/2902fb16f081

[shidel]

The short answer is mostly.

Attached was quick dump of all entries (INT, PORTS, etc) that contain a reference to SF in their ID. Skimming through it in a text editor, a lot of entries do not have information about what is used to distinguish the sub function.

However, when I added an additional filter to exclude all such SF entries which also included " subfn " in their body, the only entries which were output were from the I2C.LST. So, it appears at an initial glance the other LST files seem to likely conform to that standard.

I should also say, that the simple filter could have excluded things which do not conform. For example, if the body contained something like seealso: subfn 12h.

dump.txt.gz

[shidel]

@ecm-pushbx Here is a better dump file. It contains the Entry ID followed by all references in the entry body which mention the subfn.

It shows that items in the I2C.LST make no assertion towards any registers. However, All interrupts use that standard with either AH or AX.

dump.txt