On Monday, 96/11/25, Leslie Rhorer wrote to David Noon about "Named
Pipes" as follows:

LR>  DN> Well, I've had to resort to a little bit of assembler, since the
LR>  DN> semaphore API's seem to need 32-bit registers and 16-bit compilers
LR>  DN> won't always generate such code. I need to contact IBM via Compuserve
LR>  DN> to make sure I'm using the right calling sequence for the semaphore
LR>  DN> stuff, but we are progressing.
LR> 
LR>         This shouldn't presnt a problem.  Inline assembler 
LR> is SOP, and since all I want to do is call the assembler

Hi Les,

I've received the scoop from IBM and have the basic semaphore
interface working on my test-bed programs here. I wasn't able to use
QBASIC very readily, since CALL ABSOLUTE stinks. As a result my test
code is in C, compiled using Borland 3.1 and Watcom 10.6. The Borland
compiler often generates some bad object code when manipulating the
32-bit semaphore handles, so I am currently sticking with the 16-bit
Watcom compiler for testing.

I'll send you the assembler source for the subroutine library, C
source for the sample DOS program to be activated, and C source [unless
you'd prefer PL/I] for an OS/2 program to build the semaphore and then
activate the DOS program. You should be able to place the semaphore
requests anywhere you need them in your Power BASIC code, since the
assembler subroutines are coded using the Pascal linkage convention,
which is prettty much the lowest common denominator of such things; you
can even in-line the assembler code, I suppose. At the OS/2 end, the
semaphore API calls can replace the named pipe API calls, more or less,
in the earlier OS/2 program.

I should be done by this weekend, so check your Internet mailbox early
next week.

Regards

Dave


 * KWQ/2 1.2i * The New Chrysler Presidential Car:  The Dodge Drafter.

--- Maximus/2 3.01