/* win_ce_semaphore.c

Copyright (c) 1998, Johnson M. Hart

(with corrections 2001 by Rainer Loritz)

Permission is granted for any and all use providing that this

copyright is properly acknowledged.

There are no assurances of suitability for any use whatsoever.

WINDOWS CE: There is a collection of Windows CE functions to simulate

semaphores using only a mutex and an event. As Windows CE events cannot

be named, these simulated semaphores cannot be named either.

Implementation notes:

1. All required internal data structures are allocated on the process's heap.

2. Where appropriate, a new error code is returned (see the header

file), or, if the error is a Win32 error, that code is unchanged.

3. Notice the new handle type "SYNCHHANDLE" that has handles, counters,

and other information. This structure will grow as new objects are added

to this set; some members are specific to only one or two of the objects.

4. Mutexes are used for critical sections. These could be replaced with

CRITICAL_SECTION objects but then this would give up the time out

capability.

5. The implementation shows several interesting aspects of synchronization, some

of which are specific to Win32 and some of which are general. These are pointed

out in the comments as appropriate.

6. The wait function emulates WaitForSingleObject only. An emulation of

WaitForMultipleObjects is much harder to implement outside the kernel,

and it is not clear how to handle a mixture of WCE semaphores and normal

events and mutexes. */

#define WIN32_LEAN_AND_MEAN

#include <windows.h>

#include "win_ce_semaphore.h"

static SYNCHHANDLE CleanUp(SYNCHHANDLE hSynch, DWORD Flags);

SYNCHHANDLE

CreateSemaphoreCE(LPSECURITY_ATTRIBUTES lpSemaphoreAttributes, /* pointer to security attributes */

LONG lInitialCount, /* initial count */

LONG lMaximumCount, /* maximum count */

LPCTSTR lpName)

/* Semaphore for use with Windows CE that does not support them directly.

Requires a counter, a mutex to protect the counter, and an

autoreset event.

Here are the rules that must always hold between the autoreset event

and the mutex (any violation of these rules by the CE semaphore functions

will, in all likelihood, result in a defect):

1. No thread can set, pulse, or reset the event,

nor can it access any part of the SYNCHHANDLE structure,

without first gaining ownership of the mutex.

BUT, a thread can wait on the event without owning the mutex

(this is clearly necessary or else the event could never be set).

2. The event is in a signaled state if and only if the current semaphore

count ("CurCount") is greater than zero.

3. The semaphore count is always >= 0 and <= the maximum count */

{

SYNCHHANDLE hSynch = NULL, result = NULL;

__try {

if (lInitialCount > lMaximumCount || lMaximumCount < 0

|| lInitialCount < 0) {

/* Bad parameters */

SetLastError(SYNCH_ERROR);

__leave;

}

hSynch =

HeapAlloc(GetProcessHeap(), HEAP_ZERO_MEMORY, SYNCH_HANDLE_SIZE);

if (hSynch == NULL)

__leave;

hSynch->MaxCount = lMaximumCount;

hSynch->CurCount = lInitialCount;

hSynch->lpName = lpName;

hSynch->hMutex = CreateMutex(lpSemaphoreAttributes, FALSE, NULL);

WaitForSingleObject(hSynch->hMutex, INFINITE);

/* Create the event. It is initially signaled if and only if the

initial count is > 0 */

hSynch->hEvent = CreateEvent(lpSemaphoreAttributes, FALSE,

lInitialCount > 0, NULL);

ReleaseMutex(hSynch->hMutex);

hSynch->hSemph = NULL;

}

__finally {

/* Return with the handle, or, if there was any error, return

a null after closing any open handles and freeing any allocated memory. */

result =

CleanUp(hSynch, 6 /* An event and a mutex, but no semaphore. */ );

}

return result;

}

BOOL

ReleaseSemaphoreCE(SYNCHHANDLE hSemCE, LONG cReleaseCount,

LPLONG lpPreviousCount)

/* Windows CE equivalent to ReleaseSemaphore. */

{

BOOL Result = TRUE;

/* Gain access to the object to assure that the release count

would not cause the total count to exceed the maximum. */

__try {

WaitForSingleObject(hSemCE->hMutex, INFINITE);

/* reply only if asked to */

if (lpPreviousCount != NULL)

*lpPreviousCount = hSemCE->CurCount;

if (hSemCE->CurCount + cReleaseCount > hSemCE->MaxCount

|| cReleaseCount <= 0) {

SetLastError(SYNCH_ERROR);

Result = FALSE;

__leave;

}

hSemCE->CurCount += cReleaseCount;

/* Set the autoreset event, releasing exactly one waiting thread, now or

in the future. */

SetEvent(hSemCE->hEvent);

}

__finally {

ReleaseMutex(hSemCE->hMutex);

}

return Result;

}

DWORD

WaitForSemaphoreCE(SYNCHHANDLE hSemCE, DWORD dwMilliseconds)

/* Windows CE semaphore equivalent of WaitForSingleObject. */

{

DWORD WaitResult;

WaitResult = WaitForSingleObject(hSemCE->hMutex, dwMilliseconds);

if (WaitResult != WAIT_OBJECT_0 && WaitResult != WAIT_ABANDONED_0)

return WaitResult;

while (hSemCE->CurCount <= 0) {

/* The count is 0, and the thread must wait on the event (which, by

the rules, is currently reset) for semaphore resources to become

available. First, of course, the mutex must be released so that another

thread will be capable of setting the event. */

ReleaseMutex(hSemCE->hMutex);

/* Wait for the event to be signaled, indicating a semaphore state change.

The event is autoreset and signaled with a SetEvent (not PulseEvent)

so exactly one waiting thread (whether or not there is currently

a waiting thread) is released as a result of the SetEvent. */

WaitResult = WaitForSingleObject(hSemCE->hEvent, dwMilliseconds);

if (WaitResult != WAIT_OBJECT_0)

return WaitResult;

/* This is where the properties of setting of an autoreset event is critical

to assure that, even if the semaphore state changes between the

preceding Wait and the next, and even if NO threads are waiting

on the event at the time of the SetEvent, at least one thread

will be released.

Pulsing a manual reset event would appear to work, but it would have

a defect which could appear if the semaphore state changed between

the two waits. */

WaitResult = WaitForSingleObject(hSemCE->hMutex, dwMilliseconds);

if (WaitResult != WAIT_OBJECT_0 && WaitResult != WAIT_ABANDONED_0)

return WaitResult;

}

/* The count is not zero and this thread owns the mutex. */

hSemCE->CurCount--;

/* The event is now unsignaled, BUT, the semaphore count may not be

zero, in which case the event should be signaled again

before releasing the mutex. */

if (hSemCE->CurCount > 0)

SetEvent(hSemCE->hEvent);

ReleaseMutex(hSemCE->hMutex);

return WaitResult;

}

BOOL

CloseSynchHandle(SYNCHHANDLE hSynch)

/* Close a synchronization handle.

Improvement: Test for a valid handle before dereferencing the handle. */

{

BOOL Result = TRUE;

if (hSynch->hEvent != NULL)

Result = Result && CloseHandle(hSynch->hEvent);

if (hSynch->hMutex != NULL)

Result = Result && CloseHandle(hSynch->hMutex);

if (hSynch->hSemph != NULL)

Result = Result && CloseHandle(hSynch->hSemph);

HeapFree(GetProcessHeap(), 0, hSynch);

return (Result);

}

static SYNCHHANDLE

CleanUp(SYNCHHANDLE hSynch, DWORD Flags)

{ /* Prepare to return from a create of a synchronization handle.

If there was any failure, free any allocated resources.

"Flags" indicates which Win32 objects are required in the

synchronization handle. */

BOOL ok = TRUE;

if (hSynch == NULL)

return NULL;

if ((Flags & 4) == 1 && (hSynch->hEvent == NULL))

ok = FALSE;

if ((Flags & 2) == 1 && (hSynch->hMutex == NULL))

ok = FALSE;

if ((Flags & 1) == 1 && (hSynch->hEvent == NULL))

ok = FALSE;

if (!ok) {

CloseSynchHandle(hSynch);

return NULL;

}

/* Everything worked */

return hSynch;

}

/* vi: set ts=4 sw=4 expandtab: */