Idris2/support/refc/prim.c

268 lines
9.2 KiB
C
Raw Normal View History

#include "prim.h"
#include "refc_util.h"
RefC backend improvements (#2425) * RefC backend improvements 1. OnCollect had the wrong number of arguments. The code creator expects 3 arguments, but onCollect in prim.h expected 4 arguments. The first of which was an erased arguments. That is now fixed. 2. OnCollect did not call `newReference` when creating a new reference to the pointer and the freeing function 3. OnCollect and OnCollectAny still had a spurious printf statement Those issues have been fixed, the test case can be found in tests/refc/garbageCollect 4. The IORef mechanism expects that the %World token will be passed around consistently. This is not the case. States in Control.App make use of IORefs, but the function created from Control.App.prim_app_bind had the world token erased to NULL. Now, IORefs are managed using a global variable, IORef_Storage * global_IORef_Storage; referenced in cBackend.h, defined in the created .c file, and set to NULL in main(); 5. While multithreading and forking is still not supported, compiling a program that makes use of Control.App demands a C implementation of prim_fork. Files support/refc/threads.c and support/refc/threads.h provide a dummy implementation for it, so that Control.App programs compile and run. A test for these 2 issues is given in tests/refc/issue2424 * format changes to make the linter happy * format changes to make the linter happy * format changes to make the linter happy * spelling mistake braket -> bracket Co-authored-by: Volkmar Frinken <volkmar@onutechnology.com>
2022-04-27 15:59:32 +03:00
#include <string.h>
#include <unistd.h>
// This is NOT THREAD SAFE in the current implementation
RefC backend improvements (#2425) * RefC backend improvements 1. OnCollect had the wrong number of arguments. The code creator expects 3 arguments, but onCollect in prim.h expected 4 arguments. The first of which was an erased arguments. That is now fixed. 2. OnCollect did not call `newReference` when creating a new reference to the pointer and the freeing function 3. OnCollect and OnCollectAny still had a spurious printf statement Those issues have been fixed, the test case can be found in tests/refc/garbageCollect 4. The IORef mechanism expects that the %World token will be passed around consistently. This is not the case. States in Control.App make use of IORefs, but the function created from Control.App.prim_app_bind had the world token erased to NULL. Now, IORefs are managed using a global variable, IORef_Storage * global_IORef_Storage; referenced in cBackend.h, defined in the created .c file, and set to NULL in main(); 5. While multithreading and forking is still not supported, compiling a program that makes use of Control.App demands a C implementation of prim_fork. Files support/refc/threads.c and support/refc/threads.h provide a dummy implementation for it, so that Control.App programs compile and run. A test for these 2 issues is given in tests/refc/issue2424 * format changes to make the linter happy * format changes to make the linter happy * format changes to make the linter happy * spelling mistake braket -> bracket Co-authored-by: Volkmar Frinken <volkmar@onutechnology.com>
2022-04-27 15:59:32 +03:00
IORef_Storage *newIORef_Storage(int capacity) {
IORef_Storage *retVal = (IORef_Storage *)malloc(sizeof(IORef_Storage));
IDRIS2_REFC_VERIFY(retVal, "malloc failed");
retVal->filled = 0;
retVal->total = capacity;
retVal->refs = (Value **)malloc(sizeof(Value *) * retVal->total);
return retVal;
}
RefC backend improvements (#2425) * RefC backend improvements 1. OnCollect had the wrong number of arguments. The code creator expects 3 arguments, but onCollect in prim.h expected 4 arguments. The first of which was an erased arguments. That is now fixed. 2. OnCollect did not call `newReference` when creating a new reference to the pointer and the freeing function 3. OnCollect and OnCollectAny still had a spurious printf statement Those issues have been fixed, the test case can be found in tests/refc/garbageCollect 4. The IORef mechanism expects that the %World token will be passed around consistently. This is not the case. States in Control.App make use of IORefs, but the function created from Control.App.prim_app_bind had the world token erased to NULL. Now, IORefs are managed using a global variable, IORef_Storage * global_IORef_Storage; referenced in cBackend.h, defined in the created .c file, and set to NULL in main(); 5. While multithreading and forking is still not supported, compiling a program that makes use of Control.App demands a C implementation of prim_fork. Files support/refc/threads.c and support/refc/threads.h provide a dummy implementation for it, so that Control.App programs compile and run. A test for these 2 issues is given in tests/refc/issue2424 * format changes to make the linter happy * format changes to make the linter happy * format changes to make the linter happy * spelling mistake braket -> bracket Co-authored-by: Volkmar Frinken <volkmar@onutechnology.com>
2022-04-27 15:59:32 +03:00
void doubleIORef_Storage(IORef_Storage *ior) {
Value **values = (Value **)malloc(sizeof(Value *) * ior->total * 2);
IDRIS2_REFC_VERIFY(values, "malloc failed");
ior->total *= 2;
for (int i = 0; i < ior->filled; i++) {
values[i] = ior->refs[i];
}
free(ior->refs);
ior->refs = values;
}
Value *idris2_Data_IORef_prim__newIORef(Value *erased, Value *input_value,
Value *_world) {
RefC backend improvements (#2425) * RefC backend improvements 1. OnCollect had the wrong number of arguments. The code creator expects 3 arguments, but onCollect in prim.h expected 4 arguments. The first of which was an erased arguments. That is now fixed. 2. OnCollect did not call `newReference` when creating a new reference to the pointer and the freeing function 3. OnCollect and OnCollectAny still had a spurious printf statement Those issues have been fixed, the test case can be found in tests/refc/garbageCollect 4. The IORef mechanism expects that the %World token will be passed around consistently. This is not the case. States in Control.App make use of IORefs, but the function created from Control.App.prim_app_bind had the world token erased to NULL. Now, IORefs are managed using a global variable, IORef_Storage * global_IORef_Storage; referenced in cBackend.h, defined in the created .c file, and set to NULL in main(); 5. While multithreading and forking is still not supported, compiling a program that makes use of Control.App demands a C implementation of prim_fork. Files support/refc/threads.c and support/refc/threads.h provide a dummy implementation for it, so that Control.App programs compile and run. A test for these 2 issues is given in tests/refc/issue2424 * format changes to make the linter happy * format changes to make the linter happy * format changes to make the linter happy * spelling mistake braket -> bracket Co-authored-by: Volkmar Frinken <volkmar@onutechnology.com>
2022-04-27 15:59:32 +03:00
// if no ioRef Storag exist, start with one
if (!global_IORef_Storage) {
global_IORef_Storage = newIORef_Storage(128);
}
// expand size of needed
if (global_IORef_Storage->filled >= global_IORef_Storage->total) {
doubleIORef_Storage(global_IORef_Storage);
}
RefC backend improvements (#2425) * RefC backend improvements 1. OnCollect had the wrong number of arguments. The code creator expects 3 arguments, but onCollect in prim.h expected 4 arguments. The first of which was an erased arguments. That is now fixed. 2. OnCollect did not call `newReference` when creating a new reference to the pointer and the freeing function 3. OnCollect and OnCollectAny still had a spurious printf statement Those issues have been fixed, the test case can be found in tests/refc/garbageCollect 4. The IORef mechanism expects that the %World token will be passed around consistently. This is not the case. States in Control.App make use of IORefs, but the function created from Control.App.prim_app_bind had the world token erased to NULL. Now, IORefs are managed using a global variable, IORef_Storage * global_IORef_Storage; referenced in cBackend.h, defined in the created .c file, and set to NULL in main(); 5. While multithreading and forking is still not supported, compiling a program that makes use of Control.App demands a C implementation of prim_fork. Files support/refc/threads.c and support/refc/threads.h provide a dummy implementation for it, so that Control.App programs compile and run. A test for these 2 issues is given in tests/refc/issue2424 * format changes to make the linter happy * format changes to make the linter happy * format changes to make the linter happy * spelling mistake braket -> bracket Co-authored-by: Volkmar Frinken <volkmar@onutechnology.com>
2022-04-27 15:59:32 +03:00
// store value
Value_IORef *ioRef = IDRIS2_NEW_VALUE(Value_IORef);
ioRef->header.tag = IOREF_TAG;
ioRef->index = global_IORef_Storage->filled;
global_IORef_Storage->refs[global_IORef_Storage->filled] =
newReference(input_value);
global_IORef_Storage->filled++;
RefC backend improvements (#2425) * RefC backend improvements 1. OnCollect had the wrong number of arguments. The code creator expects 3 arguments, but onCollect in prim.h expected 4 arguments. The first of which was an erased arguments. That is now fixed. 2. OnCollect did not call `newReference` when creating a new reference to the pointer and the freeing function 3. OnCollect and OnCollectAny still had a spurious printf statement Those issues have been fixed, the test case can be found in tests/refc/garbageCollect 4. The IORef mechanism expects that the %World token will be passed around consistently. This is not the case. States in Control.App make use of IORefs, but the function created from Control.App.prim_app_bind had the world token erased to NULL. Now, IORefs are managed using a global variable, IORef_Storage * global_IORef_Storage; referenced in cBackend.h, defined in the created .c file, and set to NULL in main(); 5. While multithreading and forking is still not supported, compiling a program that makes use of Control.App demands a C implementation of prim_fork. Files support/refc/threads.c and support/refc/threads.h provide a dummy implementation for it, so that Control.App programs compile and run. A test for these 2 issues is given in tests/refc/issue2424 * format changes to make the linter happy * format changes to make the linter happy * format changes to make the linter happy * spelling mistake braket -> bracket Co-authored-by: Volkmar Frinken <volkmar@onutechnology.com>
2022-04-27 15:59:32 +03:00
return (Value *)ioRef;
}
Value *idris2_Data_IORef_prim__writeIORef(Value *erased, Value *_index,
Value *new_value, Value *_world) {
RefC backend improvements (#2425) * RefC backend improvements 1. OnCollect had the wrong number of arguments. The code creator expects 3 arguments, but onCollect in prim.h expected 4 arguments. The first of which was an erased arguments. That is now fixed. 2. OnCollect did not call `newReference` when creating a new reference to the pointer and the freeing function 3. OnCollect and OnCollectAny still had a spurious printf statement Those issues have been fixed, the test case can be found in tests/refc/garbageCollect 4. The IORef mechanism expects that the %World token will be passed around consistently. This is not the case. States in Control.App make use of IORefs, but the function created from Control.App.prim_app_bind had the world token erased to NULL. Now, IORefs are managed using a global variable, IORef_Storage * global_IORef_Storage; referenced in cBackend.h, defined in the created .c file, and set to NULL in main(); 5. While multithreading and forking is still not supported, compiling a program that makes use of Control.App demands a C implementation of prim_fork. Files support/refc/threads.c and support/refc/threads.h provide a dummy implementation for it, so that Control.App programs compile and run. A test for these 2 issues is given in tests/refc/issue2424 * format changes to make the linter happy * format changes to make the linter happy * format changes to make the linter happy * spelling mistake braket -> bracket Co-authored-by: Volkmar Frinken <volkmar@onutechnology.com>
2022-04-27 15:59:32 +03:00
Value_IORef *index = (Value_IORef *)_index;
removeReference(global_IORef_Storage->refs[index->index]);
global_IORef_Storage->refs[index->index] = newReference(new_value);
return newReference(_index);
}
// -----------------------------------
// System operations
// -----------------------------------
static Value *osstring = NULL;
Value *idris2_System_Info_prim__os(void) {
if (osstring == NULL) {
osstring = (Value *)makeString(
#ifdef _WIN32
"windows"
#elif _WIN64
"windows"
#elif __APPLE__ || __MACH__
"macOS"
#elif __linux__
"Linux"
#elif __FreeBSD__
"FreeBSD"
#elif __OpenBSD__
"OpenBSD"
#elif __NetBSD__
"NetBSD"
#elif __DragonFly__
"DragonFly"
#elif __unix || __unix__
"Unix"
#else
"Other"
#endif
);
}
return newReference(osstring);
}
// NOTE: The codegen is obviously determined at compile time,
// so the backend should optimize it by replacing it with a constant.
// It would probably also be useful for conditional compilation.
static Value *codegenstring = NULL;
Value *idris2_System_Info_prim__codegen(void) {
if (codegenstring == NULL)
codegenstring = (Value *)makeString("refc");
return newReference(codegenstring);
}
RefC backend improvements (#2425) * RefC backend improvements 1. OnCollect had the wrong number of arguments. The code creator expects 3 arguments, but onCollect in prim.h expected 4 arguments. The first of which was an erased arguments. That is now fixed. 2. OnCollect did not call `newReference` when creating a new reference to the pointer and the freeing function 3. OnCollect and OnCollectAny still had a spurious printf statement Those issues have been fixed, the test case can be found in tests/refc/garbageCollect 4. The IORef mechanism expects that the %World token will be passed around consistently. This is not the case. States in Control.App make use of IORefs, but the function created from Control.App.prim_app_bind had the world token erased to NULL. Now, IORefs are managed using a global variable, IORef_Storage * global_IORef_Storage; referenced in cBackend.h, defined in the created .c file, and set to NULL in main(); 5. While multithreading and forking is still not supported, compiling a program that makes use of Control.App demands a C implementation of prim_fork. Files support/refc/threads.c and support/refc/threads.h provide a dummy implementation for it, so that Control.App programs compile and run. A test for these 2 issues is given in tests/refc/issue2424 * format changes to make the linter happy * format changes to make the linter happy * format changes to make the linter happy * spelling mistake braket -> bracket Co-authored-by: Volkmar Frinken <volkmar@onutechnology.com>
2022-04-27 15:59:32 +03:00
Value *idris2_crash(Value *msg) {
Value_String *str = (Value_String *)msg;
printf("ERROR: %s\n", str->str);
exit(-1);
}
//
//
//
// // -----------------------------------
// // Array operations
// // -----------------------------------
Value *idris2_Data_IOArray_Prims_prim__newArray(Value *erased, Value *_length,
Value *v, Value *_word) {
RefC backend improvements (#2425) * RefC backend improvements 1. OnCollect had the wrong number of arguments. The code creator expects 3 arguments, but onCollect in prim.h expected 4 arguments. The first of which was an erased arguments. That is now fixed. 2. OnCollect did not call `newReference` when creating a new reference to the pointer and the freeing function 3. OnCollect and OnCollectAny still had a spurious printf statement Those issues have been fixed, the test case can be found in tests/refc/garbageCollect 4. The IORef mechanism expects that the %World token will be passed around consistently. This is not the case. States in Control.App make use of IORefs, but the function created from Control.App.prim_app_bind had the world token erased to NULL. Now, IORefs are managed using a global variable, IORef_Storage * global_IORef_Storage; referenced in cBackend.h, defined in the created .c file, and set to NULL in main(); 5. While multithreading and forking is still not supported, compiling a program that makes use of Control.App demands a C implementation of prim_fork. Files support/refc/threads.c and support/refc/threads.h provide a dummy implementation for it, so that Control.App programs compile and run. A test for these 2 issues is given in tests/refc/issue2424 * format changes to make the linter happy * format changes to make the linter happy * format changes to make the linter happy * spelling mistake braket -> bracket Co-authored-by: Volkmar Frinken <volkmar@onutechnology.com>
2022-04-27 15:59:32 +03:00
int length = extractInt(_length);
Value_Array *a = makeArray(length);
RefC backend improvements (#2425) * RefC backend improvements 1. OnCollect had the wrong number of arguments. The code creator expects 3 arguments, but onCollect in prim.h expected 4 arguments. The first of which was an erased arguments. That is now fixed. 2. OnCollect did not call `newReference` when creating a new reference to the pointer and the freeing function 3. OnCollect and OnCollectAny still had a spurious printf statement Those issues have been fixed, the test case can be found in tests/refc/garbageCollect 4. The IORef mechanism expects that the %World token will be passed around consistently. This is not the case. States in Control.App make use of IORefs, but the function created from Control.App.prim_app_bind had the world token erased to NULL. Now, IORefs are managed using a global variable, IORef_Storage * global_IORef_Storage; referenced in cBackend.h, defined in the created .c file, and set to NULL in main(); 5. While multithreading and forking is still not supported, compiling a program that makes use of Control.App demands a C implementation of prim_fork. Files support/refc/threads.c and support/refc/threads.h provide a dummy implementation for it, so that Control.App programs compile and run. A test for these 2 issues is given in tests/refc/issue2424 * format changes to make the linter happy * format changes to make the linter happy * format changes to make the linter happy * spelling mistake braket -> bracket Co-authored-by: Volkmar Frinken <volkmar@onutechnology.com>
2022-04-27 15:59:32 +03:00
for (int i = 0; i < length; i++) {
a->arr[i] = newReference(v);
}
RefC backend improvements (#2425) * RefC backend improvements 1. OnCollect had the wrong number of arguments. The code creator expects 3 arguments, but onCollect in prim.h expected 4 arguments. The first of which was an erased arguments. That is now fixed. 2. OnCollect did not call `newReference` when creating a new reference to the pointer and the freeing function 3. OnCollect and OnCollectAny still had a spurious printf statement Those issues have been fixed, the test case can be found in tests/refc/garbageCollect 4. The IORef mechanism expects that the %World token will be passed around consistently. This is not the case. States in Control.App make use of IORefs, but the function created from Control.App.prim_app_bind had the world token erased to NULL. Now, IORefs are managed using a global variable, IORef_Storage * global_IORef_Storage; referenced in cBackend.h, defined in the created .c file, and set to NULL in main(); 5. While multithreading and forking is still not supported, compiling a program that makes use of Control.App demands a C implementation of prim_fork. Files support/refc/threads.c and support/refc/threads.h provide a dummy implementation for it, so that Control.App programs compile and run. A test for these 2 issues is given in tests/refc/issue2424 * format changes to make the linter happy * format changes to make the linter happy * format changes to make the linter happy * spelling mistake braket -> bracket Co-authored-by: Volkmar Frinken <volkmar@onutechnology.com>
2022-04-27 15:59:32 +03:00
return (Value *)a;
}
Value *idris2_Data_IOArray_Prims_prim__arraySet(Value *erased, Value *_array,
Value *_index, Value *v,
Value *_word) {
RefC backend improvements (#2425) * RefC backend improvements 1. OnCollect had the wrong number of arguments. The code creator expects 3 arguments, but onCollect in prim.h expected 4 arguments. The first of which was an erased arguments. That is now fixed. 2. OnCollect did not call `newReference` when creating a new reference to the pointer and the freeing function 3. OnCollect and OnCollectAny still had a spurious printf statement Those issues have been fixed, the test case can be found in tests/refc/garbageCollect 4. The IORef mechanism expects that the %World token will be passed around consistently. This is not the case. States in Control.App make use of IORefs, but the function created from Control.App.prim_app_bind had the world token erased to NULL. Now, IORefs are managed using a global variable, IORef_Storage * global_IORef_Storage; referenced in cBackend.h, defined in the created .c file, and set to NULL in main(); 5. While multithreading and forking is still not supported, compiling a program that makes use of Control.App demands a C implementation of prim_fork. Files support/refc/threads.c and support/refc/threads.h provide a dummy implementation for it, so that Control.App programs compile and run. A test for these 2 issues is given in tests/refc/issue2424 * format changes to make the linter happy * format changes to make the linter happy * format changes to make the linter happy * spelling mistake braket -> bracket Co-authored-by: Volkmar Frinken <volkmar@onutechnology.com>
2022-04-27 15:59:32 +03:00
Value_Array *a = (Value_Array *)_array;
removeReference(a->arr[((Value_Int64 *)_index)->i64]);
a->arr[((Value_Int64 *)_index)->i64] = newReference(v);
return NULL;
}
//
// -----------------------------------
// Pointer operations
// -----------------------------------
Value *idris2_Prelude_IO_prim__onCollect(Value *_erased, Value *_anyPtr,
Value *_freeingFunction,
Value *_world) {
RefC backend improvements (#2425) * RefC backend improvements 1. OnCollect had the wrong number of arguments. The code creator expects 3 arguments, but onCollect in prim.h expected 4 arguments. The first of which was an erased arguments. That is now fixed. 2. OnCollect did not call `newReference` when creating a new reference to the pointer and the freeing function 3. OnCollect and OnCollectAny still had a spurious printf statement Those issues have been fixed, the test case can be found in tests/refc/garbageCollect 4. The IORef mechanism expects that the %World token will be passed around consistently. This is not the case. States in Control.App make use of IORefs, but the function created from Control.App.prim_app_bind had the world token erased to NULL. Now, IORefs are managed using a global variable, IORef_Storage * global_IORef_Storage; referenced in cBackend.h, defined in the created .c file, and set to NULL in main(); 5. While multithreading and forking is still not supported, compiling a program that makes use of Control.App demands a C implementation of prim_fork. Files support/refc/threads.c and support/refc/threads.h provide a dummy implementation for it, so that Control.App programs compile and run. A test for these 2 issues is given in tests/refc/issue2424 * format changes to make the linter happy * format changes to make the linter happy * format changes to make the linter happy * spelling mistake braket -> bracket Co-authored-by: Volkmar Frinken <volkmar@onutechnology.com>
2022-04-27 15:59:32 +03:00
Value_GCPointer *retVal = IDRIS2_NEW_VALUE(Value_GCPointer);
retVal->header.tag = GC_POINTER_TAG;
retVal->p = (Value_Pointer *)newReference(_anyPtr);
retVal->onCollectFct = (Value_Closure *)newReference(_freeingFunction);
return (Value *)retVal;
}
Value *idris2_Prelude_IO_prim__onCollectAny(Value *_anyPtr,
Value *_freeingFunction,
Value *_world) {
RefC backend improvements (#2425) * RefC backend improvements 1. OnCollect had the wrong number of arguments. The code creator expects 3 arguments, but onCollect in prim.h expected 4 arguments. The first of which was an erased arguments. That is now fixed. 2. OnCollect did not call `newReference` when creating a new reference to the pointer and the freeing function 3. OnCollect and OnCollectAny still had a spurious printf statement Those issues have been fixed, the test case can be found in tests/refc/garbageCollect 4. The IORef mechanism expects that the %World token will be passed around consistently. This is not the case. States in Control.App make use of IORefs, but the function created from Control.App.prim_app_bind had the world token erased to NULL. Now, IORefs are managed using a global variable, IORef_Storage * global_IORef_Storage; referenced in cBackend.h, defined in the created .c file, and set to NULL in main(); 5. While multithreading and forking is still not supported, compiling a program that makes use of Control.App demands a C implementation of prim_fork. Files support/refc/threads.c and support/refc/threads.h provide a dummy implementation for it, so that Control.App programs compile and run. A test for these 2 issues is given in tests/refc/issue2424 * format changes to make the linter happy * format changes to make the linter happy * format changes to make the linter happy * spelling mistake braket -> bracket Co-authored-by: Volkmar Frinken <volkmar@onutechnology.com>
2022-04-27 15:59:32 +03:00
Value_GCPointer *retVal = IDRIS2_NEW_VALUE(Value_GCPointer);
retVal->header.tag = GC_POINTER_TAG;
retVal->p = (Value_Pointer *)newReference(_anyPtr);
retVal->onCollectFct = (Value_Closure *)newReference(_freeingFunction);
return (Value *)retVal;
}
// Threads
// %foreign "scheme:blodwen-mutex"
// prim__makeMutex : PrimIO Mutex
RefC backend improvements (#2425) * RefC backend improvements 1. OnCollect had the wrong number of arguments. The code creator expects 3 arguments, but onCollect in prim.h expected 4 arguments. The first of which was an erased arguments. That is now fixed. 2. OnCollect did not call `newReference` when creating a new reference to the pointer and the freeing function 3. OnCollect and OnCollectAny still had a spurious printf statement Those issues have been fixed, the test case can be found in tests/refc/garbageCollect 4. The IORef mechanism expects that the %World token will be passed around consistently. This is not the case. States in Control.App make use of IORefs, but the function created from Control.App.prim_app_bind had the world token erased to NULL. Now, IORefs are managed using a global variable, IORef_Storage * global_IORef_Storage; referenced in cBackend.h, defined in the created .c file, and set to NULL in main(); 5. While multithreading and forking is still not supported, compiling a program that makes use of Control.App demands a C implementation of prim_fork. Files support/refc/threads.c and support/refc/threads.h provide a dummy implementation for it, so that Control.App programs compile and run. A test for these 2 issues is given in tests/refc/issue2424 * format changes to make the linter happy * format changes to make the linter happy * format changes to make the linter happy * spelling mistake braket -> bracket Co-authored-by: Volkmar Frinken <volkmar@onutechnology.com>
2022-04-27 15:59:32 +03:00
// using pthread_mutex_init(pthread_mutex_t *mutex, const pthread_mutexattr_t
// *attr)
Value *System_Concurrency_Raw_prim__makeMutex(Value *_world) {
Value_Mutex *mut = IDRIS2_NEW_VALUE(Value_Mutex);
mut->header.tag = MUTEX_TAG;
mut->mutex = (pthread_mutex_t *)malloc(sizeof(pthread_mutex_t));
int r = pthread_mutex_init(mut->mutex, NULL);
IDRIS2_REFC_VERIFY(!r, "pthread_mutex_init failed: %s", strerror(r));
return (Value *)mut;
}
// %foreign "scheme:blodwen-lock"
// prim__mutexAcquire : Mutex -> PrimIO ()
// using pthread_mutex_lock(pthread_mutex_t *mutex)
RefC backend improvements (#2425) * RefC backend improvements 1. OnCollect had the wrong number of arguments. The code creator expects 3 arguments, but onCollect in prim.h expected 4 arguments. The first of which was an erased arguments. That is now fixed. 2. OnCollect did not call `newReference` when creating a new reference to the pointer and the freeing function 3. OnCollect and OnCollectAny still had a spurious printf statement Those issues have been fixed, the test case can be found in tests/refc/garbageCollect 4. The IORef mechanism expects that the %World token will be passed around consistently. This is not the case. States in Control.App make use of IORefs, but the function created from Control.App.prim_app_bind had the world token erased to NULL. Now, IORefs are managed using a global variable, IORef_Storage * global_IORef_Storage; referenced in cBackend.h, defined in the created .c file, and set to NULL in main(); 5. While multithreading and forking is still not supported, compiling a program that makes use of Control.App demands a C implementation of prim_fork. Files support/refc/threads.c and support/refc/threads.h provide a dummy implementation for it, so that Control.App programs compile and run. A test for these 2 issues is given in tests/refc/issue2424 * format changes to make the linter happy * format changes to make the linter happy * format changes to make the linter happy * spelling mistake braket -> bracket Co-authored-by: Volkmar Frinken <volkmar@onutechnology.com>
2022-04-27 15:59:32 +03:00
Value *System_Concurrency_Raw_prim__mutexAcquire(Value *_mutex, Value *_world) {
int r = pthread_mutex_lock(((Value_Mutex *)_mutex)->mutex);
IDRIS2_REFC_VERIFY(!r, "pthread_mutex_lock failed: %s", strerror(r));
return NULL;
}
// %foreign "scheme:blodwen-unlock"
// prim__mutexRelease : Mutex -> PrimIO ()
RefC backend improvements (#2425) * RefC backend improvements 1. OnCollect had the wrong number of arguments. The code creator expects 3 arguments, but onCollect in prim.h expected 4 arguments. The first of which was an erased arguments. That is now fixed. 2. OnCollect did not call `newReference` when creating a new reference to the pointer and the freeing function 3. OnCollect and OnCollectAny still had a spurious printf statement Those issues have been fixed, the test case can be found in tests/refc/garbageCollect 4. The IORef mechanism expects that the %World token will be passed around consistently. This is not the case. States in Control.App make use of IORefs, but the function created from Control.App.prim_app_bind had the world token erased to NULL. Now, IORefs are managed using a global variable, IORef_Storage * global_IORef_Storage; referenced in cBackend.h, defined in the created .c file, and set to NULL in main(); 5. While multithreading and forking is still not supported, compiling a program that makes use of Control.App demands a C implementation of prim_fork. Files support/refc/threads.c and support/refc/threads.h provide a dummy implementation for it, so that Control.App programs compile and run. A test for these 2 issues is given in tests/refc/issue2424 * format changes to make the linter happy * format changes to make the linter happy * format changes to make the linter happy * spelling mistake braket -> bracket Co-authored-by: Volkmar Frinken <volkmar@onutechnology.com>
2022-04-27 15:59:32 +03:00
// using int pthread_mutex_unlock(pthread_mutex_t *mutex)
Value *System_Concurrency_Raw_prim__mutexRelease(Value *_mutex, Value *_world) {
int r = pthread_mutex_unlock(((Value_Mutex *)_mutex)->mutex);
IDRIS2_REFC_VERIFY(!r, "pthread_mutex_unlock failed: %s", strerror(r));
return NULL;
}
// %foreign "scheme:blodwen-condition"
// prim__makeCondition : PrimIO Condition
RefC backend improvements (#2425) * RefC backend improvements 1. OnCollect had the wrong number of arguments. The code creator expects 3 arguments, but onCollect in prim.h expected 4 arguments. The first of which was an erased arguments. That is now fixed. 2. OnCollect did not call `newReference` when creating a new reference to the pointer and the freeing function 3. OnCollect and OnCollectAny still had a spurious printf statement Those issues have been fixed, the test case can be found in tests/refc/garbageCollect 4. The IORef mechanism expects that the %World token will be passed around consistently. This is not the case. States in Control.App make use of IORefs, but the function created from Control.App.prim_app_bind had the world token erased to NULL. Now, IORefs are managed using a global variable, IORef_Storage * global_IORef_Storage; referenced in cBackend.h, defined in the created .c file, and set to NULL in main(); 5. While multithreading and forking is still not supported, compiling a program that makes use of Control.App demands a C implementation of prim_fork. Files support/refc/threads.c and support/refc/threads.h provide a dummy implementation for it, so that Control.App programs compile and run. A test for these 2 issues is given in tests/refc/issue2424 * format changes to make the linter happy * format changes to make the linter happy * format changes to make the linter happy * spelling mistake braket -> bracket Co-authored-by: Volkmar Frinken <volkmar@onutechnology.com>
2022-04-27 15:59:32 +03:00
// using int pthread_cond_init(pthread_cond_t *cond, const pthread_condattr_t
// *attr) with standard initialisation
Value *System_Concurrency_Raw_prim__makeCondition(Value *_world) {
// typedef struct{
// Value_header header;
// pthread_cond_t *cond;
// }Value_Condition;
Value_Condition *c = IDRIS2_NEW_VALUE(Value_Condition);
c->header.tag = CONDITION_TAG;
c->cond = (pthread_cond_t *)malloc(sizeof(pthread_cond_t));
IDRIS2_REFC_VERIFY(c->cond, "malloc failed");
int r = pthread_cond_init(c->cond, NULL);
IDRIS2_REFC_VERIFY(!r, "pthread_cond_init failed: %s", strerror(r));
return (Value *)c;
}
// %foreign "scheme:blodwen-condition-wait"
// prim__conditionWait : Condition -> Mutex -> PrimIO ()
// using int pthread_cond_wait(pthread_cond_t *, pthread_mutex_t *mutex)
RefC backend improvements (#2425) * RefC backend improvements 1. OnCollect had the wrong number of arguments. The code creator expects 3 arguments, but onCollect in prim.h expected 4 arguments. The first of which was an erased arguments. That is now fixed. 2. OnCollect did not call `newReference` when creating a new reference to the pointer and the freeing function 3. OnCollect and OnCollectAny still had a spurious printf statement Those issues have been fixed, the test case can be found in tests/refc/garbageCollect 4. The IORef mechanism expects that the %World token will be passed around consistently. This is not the case. States in Control.App make use of IORefs, but the function created from Control.App.prim_app_bind had the world token erased to NULL. Now, IORefs are managed using a global variable, IORef_Storage * global_IORef_Storage; referenced in cBackend.h, defined in the created .c file, and set to NULL in main(); 5. While multithreading and forking is still not supported, compiling a program that makes use of Control.App demands a C implementation of prim_fork. Files support/refc/threads.c and support/refc/threads.h provide a dummy implementation for it, so that Control.App programs compile and run. A test for these 2 issues is given in tests/refc/issue2424 * format changes to make the linter happy * format changes to make the linter happy * format changes to make the linter happy * spelling mistake braket -> bracket Co-authored-by: Volkmar Frinken <volkmar@onutechnology.com>
2022-04-27 15:59:32 +03:00
Value *System_Concurrency_Raw_prim__conditionWait(Value *_condition,
Value *_mutex,
Value *_world) {
Value_Condition *cond = (Value_Condition *)_condition;
Value_Mutex *mutex = (Value_Mutex *)_mutex;
int r = pthread_cond_wait(cond->cond, mutex->mutex);
IDRIS2_REFC_VERIFY(!r, "pthread_cond_wait failed: %s", strerror(r));
return NULL;
}
// %foreign "scheme:blodwen-condition-wait-timeout"
// prim__conditionWaitTimeout : Condition -> Mutex -> Int -> PrimIO ()
RefC backend improvements (#2425) * RefC backend improvements 1. OnCollect had the wrong number of arguments. The code creator expects 3 arguments, but onCollect in prim.h expected 4 arguments. The first of which was an erased arguments. That is now fixed. 2. OnCollect did not call `newReference` when creating a new reference to the pointer and the freeing function 3. OnCollect and OnCollectAny still had a spurious printf statement Those issues have been fixed, the test case can be found in tests/refc/garbageCollect 4. The IORef mechanism expects that the %World token will be passed around consistently. This is not the case. States in Control.App make use of IORefs, but the function created from Control.App.prim_app_bind had the world token erased to NULL. Now, IORefs are managed using a global variable, IORef_Storage * global_IORef_Storage; referenced in cBackend.h, defined in the created .c file, and set to NULL in main(); 5. While multithreading and forking is still not supported, compiling a program that makes use of Control.App demands a C implementation of prim_fork. Files support/refc/threads.c and support/refc/threads.h provide a dummy implementation for it, so that Control.App programs compile and run. A test for these 2 issues is given in tests/refc/issue2424 * format changes to make the linter happy * format changes to make the linter happy * format changes to make the linter happy * spelling mistake braket -> bracket Co-authored-by: Volkmar Frinken <volkmar@onutechnology.com>
2022-04-27 15:59:32 +03:00
// using int pthread_cond_timedwait(pthread_cond_t *cond, pthread_mutex_t
// *mutex, const struct timespec *abstime)
Value *System_Concurrency_Raw_prim__conditionWaitTimeout(Value *_condition,
Value *_mutex,
Value *_timeout,
Value *_world) {
Value_Condition *cond = (Value_Condition *)_condition;
Value_Mutex *mutex = (Value_Mutex *)_mutex;
Value_Int64 *timeout = (Value_Int64 *)_timeout;
struct timespec t;
t.tv_sec = timeout->i64 / 1000000;
t.tv_nsec = timeout->i64 % 1000000;
int r = pthread_cond_timedwait(cond->cond, mutex->mutex, &t);
IDRIS2_REFC_VERIFY(!r, "pthread_cond_timedwait failed: %s", strerror(r));
return NULL;
}
// %foreign "scheme:blodwen-condition-signal"
// prim__conditionSignal : Condition -> PrimIO ()
// using int pthread_cond_signal(pthread_cond_t *cond)
RefC backend improvements (#2425) * RefC backend improvements 1. OnCollect had the wrong number of arguments. The code creator expects 3 arguments, but onCollect in prim.h expected 4 arguments. The first of which was an erased arguments. That is now fixed. 2. OnCollect did not call `newReference` when creating a new reference to the pointer and the freeing function 3. OnCollect and OnCollectAny still had a spurious printf statement Those issues have been fixed, the test case can be found in tests/refc/garbageCollect 4. The IORef mechanism expects that the %World token will be passed around consistently. This is not the case. States in Control.App make use of IORefs, but the function created from Control.App.prim_app_bind had the world token erased to NULL. Now, IORefs are managed using a global variable, IORef_Storage * global_IORef_Storage; referenced in cBackend.h, defined in the created .c file, and set to NULL in main(); 5. While multithreading and forking is still not supported, compiling a program that makes use of Control.App demands a C implementation of prim_fork. Files support/refc/threads.c and support/refc/threads.h provide a dummy implementation for it, so that Control.App programs compile and run. A test for these 2 issues is given in tests/refc/issue2424 * format changes to make the linter happy * format changes to make the linter happy * format changes to make the linter happy * spelling mistake braket -> bracket Co-authored-by: Volkmar Frinken <volkmar@onutechnology.com>
2022-04-27 15:59:32 +03:00
Value *System_Concurrency_Raw_prim__conditionSignal(Value *_condition,
Value *_world) {
Value_Condition *cond = (Value_Condition *)_condition;
int r = pthread_cond_signal(cond->cond);
IDRIS2_REFC_VERIFY(!r, "pthread_cond_signal failed: %s", strerror(r));
return NULL;
}
// %foreign "scheme:blodwen-condition-broadcast"
// prim__conditionBroadcast : Condition -> PrimIO ()
// using int pthread_cond_broadcast(pthread_cond_t *cond)
RefC backend improvements (#2425) * RefC backend improvements 1. OnCollect had the wrong number of arguments. The code creator expects 3 arguments, but onCollect in prim.h expected 4 arguments. The first of which was an erased arguments. That is now fixed. 2. OnCollect did not call `newReference` when creating a new reference to the pointer and the freeing function 3. OnCollect and OnCollectAny still had a spurious printf statement Those issues have been fixed, the test case can be found in tests/refc/garbageCollect 4. The IORef mechanism expects that the %World token will be passed around consistently. This is not the case. States in Control.App make use of IORefs, but the function created from Control.App.prim_app_bind had the world token erased to NULL. Now, IORefs are managed using a global variable, IORef_Storage * global_IORef_Storage; referenced in cBackend.h, defined in the created .c file, and set to NULL in main(); 5. While multithreading and forking is still not supported, compiling a program that makes use of Control.App demands a C implementation of prim_fork. Files support/refc/threads.c and support/refc/threads.h provide a dummy implementation for it, so that Control.App programs compile and run. A test for these 2 issues is given in tests/refc/issue2424 * format changes to make the linter happy * format changes to make the linter happy * format changes to make the linter happy * spelling mistake braket -> bracket Co-authored-by: Volkmar Frinken <volkmar@onutechnology.com>
2022-04-27 15:59:32 +03:00
Value *System_Concurrency_Raw_prim__conditionBroadcast(Value *_condition,
Value *_mutex) {
Value_Condition *cond = (Value_Condition *)_condition;
int r = pthread_cond_broadcast(cond->cond);
IDRIS2_REFC_VERIFY(!r, "pthread_cond_broadcast failed: %s", strerror(r));
return NULL;
}