msm: kgsl: Affine kgsl_3d0_irq and worker kthread to the big CPU cluster

These are in the critical path for rendering frames to the display, so
mark them as performance-critical and affine them to the big CPU
cluster. They aren't placed onto the prime cluster because the
single-CPU prime cluster will be used to run the DRM IRQ and kthreads.
DRM is more latency-critical than KGSL and we need to have DRM and KGSL
running on separate CPUs for the best performance, so KGSL gets the big
cluster.

Note that since there are other IRQs requested via kgsl_request_irq(),
we must specify that the IRQ to be made perf-critical is kgsl_3d0_irq.

Change-Id: Ia86a553e41632789aae5bf8ebaeb7d969b67e4ae
Signed-off-by: wHo-EM-i <ehteshammalik18998@gmail.com>
Signed-off-by: Santhosh <santhosh.user.why.red@gmail.com>

drivers/gpu/msm/kgsl.c

@@ -5283,12 +5283,15 @@ int kgsl_request_irq(struct platform_device *pdev, const  char *name,
 		irq_handler_t handler, void *data)
 {
 	int ret, num = platform_get_irq_byname(pdev, name);
+	unsigned long irqflags = IRQF_TRIGGER_HIGH;
 
 	if (num < 0)
 		return num;
 
-	ret = devm_request_irq(&pdev->dev, num, handler, IRQF_TRIGGER_HIGH,
-		name, data);
+	if (!strcmp(name, "kgsl_3d0_irq"))
+		irqflags |= IRQF_PERF_AFFINE;
+
+	ret = devm_request_irq(&pdev->dev, num, handler, irqflags, name, data);
 
 	if (ret)
 		dev_err(&pdev->dev, "Unable to get interrupt %s: %d\n",
@@ -5574,8 +5577,8 @@ static int __init kgsl_core_init(void)
 
 	kthread_init_worker(&kgsl_driver.worker);
 
-	kgsl_driver.worker_thread = kthread_run(kthread_worker_fn,
-		&kgsl_driver.worker, "kgsl_worker_thread");
+	kgsl_driver.worker_thread = kthread_run_perf_critical(cpu_perf_mask,
+		kthread_worker_fn, &kgsl_driver.worker, "kgsl_worker_thread");
 
 	if (IS_ERR(kgsl_driver.worker_thread)) {
 		pr_err("kgsl: unable to start kgsl thread\n");