drm/amdkfd: Track when module's init is complete

Current dependencies between amdkfd and radeon/amdgpu force the loading
of amdkfd _before_ radeon and/or amdgpu are loaded. When all these kernel
drivers are built as modules, this ordering is enforced by the kernel
built-in mechanism of loading dependent modules.

However, there is no such mechanism in case where all these drivers are
compiled inside the kernel image (not as modules). The current way to
enforce loading of amdkfd before radeon/amdgpu, is to put amdkfd before
radeon/amdgpu in the drm Makefile, but that method is way too fragile.

In addition, there is no kernel mechanism to check whether a kernel
driver that is built inside the kernel image, has already been loaded.

To solve this, this patch adds to kfd_module.c a new static variable,
amdkfd_init_completed, that is set to 1 only when amdkfd's
module initialization function has been completed (successfully).

kgd2kfd_init(), which is the initialization function of the
kgd-->kfd interface, and which is the first function in amdkfd called by
radeon/amdgpu, will return successfully only if amdkfd_init_completed is
equal 1.

If amdkfd_init_completed is not equal to 1, kgd2kfd_init() will
return -EPROBE_DEFER to signal radeon/amdgpu they need to defer
their loading until amdkfd is loaded.

Signed-off-by: Oded Gabbay <oded.gabbay@gmail.com>
Reviewed-by: Alex Deucher <alexander.deucher@amd.com>

drivers/gpu/drm/amd/amdgpu/amdgpu_amdkfd.c

@@ -70,7 +70,7 @@ bool amdgpu_amdkfd_load_interface(struct amdgpu_device *rdev)
 		return false;
 	}
 
-	if (!kgd2kfd_init_p(KFD_INTERFACE_VERSION, &kgd2kfd)) {
+	if (kgd2kfd_init_p(KFD_INTERFACE_VERSION, &kgd2kfd)) {
 		symbol_put(kgd2kfd_init);
 		kfd2kgd = NULL;
 		kgd2kfd = NULL;
@@ -80,7 +80,7 @@ bool amdgpu_amdkfd_load_interface(struct amdgpu_device *rdev)
 
 	return true;
 #elif defined(CONFIG_HSA_AMD)
-	if (!kgd2kfd_init(KFD_INTERFACE_VERSION, &kgd2kfd)) {
+	if (kgd2kfd_init(KFD_INTERFACE_VERSION, &kgd2kfd)) {
 		kfd2kgd = NULL;
 		kgd2kfd = NULL;
 		return false;

drivers/gpu/drm/amd/amdkfd/kfd_module.c

@@ -59,18 +59,23 @@ module_param(send_sigterm, int, 0444);
 MODULE_PARM_DESC(send_sigterm,
 	"Send sigterm to HSA process on unhandled exception (0 = disable, 1 = enable)");
 
-bool kgd2kfd_init(unsigned interface_version, const struct kgd2kfd_calls **g2f)
+static int amdkfd_init_completed;
+
+int kgd2kfd_init(unsigned interface_version, const struct kgd2kfd_calls **g2f)
 {
+	if (!amdkfd_init_completed)
+		return -EPROBE_DEFER;
+
 	/*
 	 * Only one interface version is supported,
 	 * no kfd/kgd version skew allowed.
 	 */
 	if (interface_version != KFD_INTERFACE_VERSION)
-		return false;
+		return -EINVAL;
 
 	*g2f = &kgd2kfd;
 
-	return true;
+	return 0;
 }
 EXPORT_SYMBOL(kgd2kfd_init);
 
@@ -111,6 +116,8 @@ static int __init kfd_module_init(void)
 
 	kfd_process_create_wq();
 
+	amdkfd_init_completed = 1;
+
 	dev_info(kfd_device, "Initialized module\n");
 
 	return 0;
@@ -125,6 +132,8 @@ err_pasid:
 
 static void __exit kfd_module_exit(void)
 {
+	amdkfd_init_completed = 0;
+
 	kfd_process_destroy_wq();
 	kfd_topology_shutdown();
 	kfd_chardev_exit();

drivers/gpu/drm/amd/include/kgd_kfd_interface.h

@@ -221,7 +221,7 @@ struct kgd2kfd_calls {
 	int (*resume)(struct kfd_dev *kfd);
 };
 
-bool kgd2kfd_init(unsigned interface_version,
+int kgd2kfd_init(unsigned interface_version,
 		const struct kgd2kfd_calls **g2f);
 
 #endif	/* KGD_KFD_INTERFACE_H_INCLUDED */

drivers/gpu/drm/radeon/radeon_kfd.c

@@ -142,7 +142,7 @@ bool radeon_kfd_init(void)
 	if (kgd2kfd_init_p == NULL)
 		return false;
 
-	if (!kgd2kfd_init_p(KFD_INTERFACE_VERSION, &kgd2kfd)) {
+	if (kgd2kfd_init_p(KFD_INTERFACE_VERSION, &kgd2kfd)) {
 		symbol_put(kgd2kfd_init);
 		kgd2kfd = NULL;
 
@@ -151,7 +151,7 @@ bool radeon_kfd_init(void)
 
 	return true;
 #elif defined(CONFIG_HSA_AMD)
-	if (!kgd2kfd_init(KFD_INTERFACE_VERSION, &kgd2kfd)) {
+	if (kgd2kfd_init(KFD_INTERFACE_VERSION, &kgd2kfd)) {
 		kgd2kfd = NULL;
 
 		return false;