c to asm

[Jimg]

Would someone convert this code to masm for me?

http://msdn.microsoft.com/en-us/library/ms713937(VS.85).aspx

Code Select Expand

void CALLBACK OneShotTimer(UINT wTimerID, UINT msg,
    DWORD dwUser, DWORD dw1, DWORD dw2)
{
    NPSEQ npSeq;             // pointer to sequencer data
    npSeq = (NPSEQ)dwUser;
    npSeq->wTimerID = 0;     // invalidate timer ID (no longer in use)
    TimerRoutine(npSeq);     // handle tasks
}
This is even more cryptic than your normal c code.

[Infro_X]

Code Select Expand


OneShotTimer:
mov eax,[esp+0Ch]
mov WORD PTR [eax + SEQ.wTimerID],0
invoke TimerRoutine,eax
retn 14h


[ToutEnMasm]

Hello,
And if you Have other questions like that,add the option "/FA" (like Flat Assembler) to the C compiler and .... you have an .asm file.

[Jimg]

Thanks Infro_X.

ToutEnMasm-  I don't have a C compiler, and I can just barely read and guess at what half the code I see is doing.

I take that back.  I bought Quick-C about 20 years ago, just to use the Quick-Assembler option.  Not C++ at all.
Yup, I just looked at the books.  Copyright 1988.

[mitchi]

Code Select Expand


OneShotTimer:
mov eax,[esp+0Ch]
mov WORD PTR [eax + SEQ.wTimerID],0
invoke TimerRoutine,eax
retn 14h


Is it not this instead :
mov eax,[esp+16]

First argument is ESP+8
Next is ESP +12
Third is ESP + 16

[MichaelW]

After spending about an hour searching, I can't find a declaration for NPSEQ and the structure it points to, and without that I can't see any way to actually use the code.

[Jimg]

Thank you Michael, that was my experience also.  It was never my intention anyone waste an hour, even though I have wasted ten trying to get a multimedia timer that is more accurate than a simple timer.  I no longer need it urgently, but I hope to get back to it some time.

[MichaelW]

I spent an hour on it mostly because I'm stubborn. In the Windows Server 2003 PSDK there are three instances of NPSEQ, including one in a copy of the code you posted, so it must be defined in a header file somewhere. A search on NPSEQ turns up pages that do not appear to contain NPSEQ or anything related, so I'm guessing that it's so frequently searched for that it's being inserted somehow to fool the search engines.

[jj2007]

It seems it's simply a pointer to a user-defined data structure. And it's oooold - this is from Win 3.1 SDK:

This section describes how an application might use the timer services. First, the application calls the timeGetDevCaps function to determine the minimum and maximum resolution supported by the timer services. Before setting up any timer events, the application uses timeBeginPeriod to establish the minimum timer resolution it will use, as shown in the following code fragment:
 
#define TARGET_RESOLUTION 1         // Try for 1-millisecond accuracy

TIMECAPS tc;
UINT     wTimerRes;

if(timeGetDevCaps(&tc, sizeof(TIMECAPS)) != TIMERR_NOERROR)
{
    // Error; application can't continue
}

wTimerRes = min(max(tc.wPeriodMin, TARGET_RESOLUTION), tc.wPeriodMax);
timeBeginPeriod(wTimerRes);
 
To start the timer event, the application specifies the amount of time before the callback occurs, the required resolution, the address of the callback function, and user data to supply with the callback. The application might use a function like the following to start a one-time timer event:
 
UINT SetTimerCallback(NPSEQ npSeq,          // Sequencer data
                      UINT msInterval)      // Event interval
{
    npSeq->wTimerID = timeSetEvent(
         msInterval,                        // Delay
         wTimerRes,                         // Resolution (global variable)
         OneShotCallback,                   // Callback function
         (DWORD)npSeq,                      // User data
         TIME_ONESHOT );                    // Event type (one-time)

    if(! npSeq->wTimerID)
        return ERR_TIMER;
    else
        return ERR_NOERROR;
}
 
The following callback function resides in a fixed code segment in a DLL. It is limited to calling those functions that are interrupt-callable. The TimerIntRoutine procedure it calls also resides in a fixed code segment.
 
void FAR PASCAL
OneShotTimer(UINT wTimerID, UINT msg, DWORD dwUser, DWORD dw1, DWORD dw2)
{
    NPSEQ npSeq;              // Pointer to sequencer data

    npSeq = (NPSEQ)dwUser;
    npSeq->wTimerID = 0;      // Invalidate timer id, since no longer in use

    TimerIntRoutine(npSeq);   // Handle interrupt-time tasks
}
 
Before freeing the DLL that contains the callback function, the application cancels any outstanding timers. To cancel one timer event, it might call the following function:
 
void DestroyTimer(NPSEQ npSeq)
{
    if(npSeq->wTimerID)                     // If timer event is pending
    {
        timeKillEvent(npSeq->wTimerID);     // Cancel the event
        npSeq->wTimerID = 0;
    }
}
 
Finally, to cancel the minimum timer resolution it established, the application calls timeEndPeriod as follows:
 
timeEndPeriod(wTimerRes);

[Jimg]

Nice.  How long before you convert that to Masm?

[mitchi]

How about learning some C ?
Plain C is so easy...

[Jimg]

Irony aside, if it seems clean and straightforward to you, then there is clearly no reason for me to say anything further.

[mitchi]

I learned assembly language studying the assembly output of my C programs. I'm a special case I guess.

[redskull]

I learned assembly language studying the assembly output of my C programs. I'm a special case I guess.

Ironic; I actually learned C by studying assembly.  Basically, i was looking for something that could do low level Win32 API, but retain the high level macro features of MASM, and automatically do array scaling and such; it's actually a little scary how similar low level C and MASM w/ high-level macros are to each other...

-ac

[xanatose]

I learned assembly language studying the assembly output of my C programs. I'm a special case I guess.

I did the same, back on the days of turbo C, 3.0