Merge tag 'amd-drm-next-6.12-2024-09-13' of https://gitlab.freedesktop.org/agd5f/linux into drm-next

amd-drm-next-6.12-2024-09-13:

amdgpu:
- GPUVM sync fixes
- kdoc fixes
- Misc spelling mistakes
- Add some raven GFXOFF quirks
- Use clamp helper
- DC fixes
- JPEG fixes
- Process isolation fix
- Queue reset fix
- W=1 cleanup
- SMU14 fixes
- JPEG fixes

amdkfd:
- Fetch cacheline info from IP discovery
- Queue reset fix
- RAS fix
- Document SVM events
- CRIU fixes
- Race fix in dma-buf handling

drm:
- dma-buf fd race fixes

Signed-off-by: Dave Airlie <[email protected]>

From: Alex Deucher <[email protected]>
Link: https://patchwork.freedesktop.org/patch/msgid/[email protected]

Author: airlied

drivers/gpu/drm/amd/amdgpu/Makefile

@@ -39,23 +39,7 @@ ccflags-y := -I$(FULL_AMD_PATH)/include/asic_reg \
 	-I$(FULL_AMD_DISPLAY_PATH)/amdgpu_dm \
 	-I$(FULL_AMD_PATH)/amdkfd
 
-subdir-ccflags-y := -Wextra
-subdir-ccflags-y += -Wunused
-subdir-ccflags-y += -Wmissing-prototypes
-subdir-ccflags-y += -Wmissing-declarations
-subdir-ccflags-y += -Wmissing-include-dirs
-subdir-ccflags-y += -Wold-style-definition
-subdir-ccflags-y += -Wmissing-format-attribute
-# Need this to avoid recursive variable evaluation issues
-cond-flags := $(call cc-option, -Wunused-but-set-variable) \
-	$(call cc-option, -Wunused-const-variable) \
-	$(call cc-option, -Wstringop-truncation) \
-	$(call cc-option, -Wpacked-not-aligned)
-subdir-ccflags-y += $(cond-flags)
-subdir-ccflags-y += -Wno-unused-parameter
-subdir-ccflags-y += -Wno-type-limits
-subdir-ccflags-y += -Wno-sign-compare
-subdir-ccflags-y += -Wno-missing-field-initializers
+# Locally disable W=1 warnings enabled in drm subsystem Makefile
 subdir-ccflags-y += -Wno-override-init
 subdir-ccflags-$(CONFIG_DRM_AMDGPU_WERROR) += -Werror
 

drivers/gpu/drm/amd/amdgpu/amdgpu.h

@@ -237,6 +237,7 @@ extern int sched_policy;
 extern bool debug_evictions;
 extern bool no_system_mem_limit;
 extern int halt_if_hws_hang;
+extern uint amdgpu_svm_default_granularity;
 #else
 static const int __maybe_unused sched_policy = KFD_SCHED_POLICY_HWS;
 static const bool __maybe_unused debug_evictions; /* = false */

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

@@ -20,7 +20,6 @@
  * OTHER DEALINGS IN THE SOFTWARE.
  */
 #include <linux/module.h>
-#include <linux/fdtable.h>
 #include <linux/uaccess.h>
 #include <linux/firmware.h>
 #include "amdgpu.h"

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

@@ -25,7 +25,6 @@
 #include <linux/pagemap.h>
 #include <linux/sched/mm.h>
 #include <linux/sched/task.h>
-#include <linux/fdtable.h>
 #include <drm/ttm/ttm_tt.h>
 
 #include <drm/drm_exec.h>
@@ -818,18 +817,13 @@ static int kfd_mem_export_dmabuf(struct kgd_mem *mem)
 	if (!mem->dmabuf) {
 		struct amdgpu_device *bo_adev;
 		struct dma_buf *dmabuf;
-		int r, fd;
 
 		bo_adev = amdgpu_ttm_adev(mem->bo->tbo.bdev);
-		r = drm_gem_prime_handle_to_fd(&bo_adev->ddev, bo_adev->kfd.client.file,
+		dmabuf = drm_gem_prime_handle_to_dmabuf(&bo_adev->ddev, bo_adev->kfd.client.file,
 					       mem->gem_handle,
 			mem->alloc_flags & KFD_IOC_ALLOC_MEM_FLAGS_WRITABLE ?
-					       DRM_RDWR : 0, &fd);
-		if (r)
-			return r;
-		dmabuf = dma_buf_get(fd);
-		close_fd(fd);
-		if (WARN_ON_ONCE(IS_ERR(dmabuf)))
+					       DRM_RDWR : 0);
+		if (IS_ERR(dmabuf))
 			return PTR_ERR(dmabuf);
 		mem->dmabuf = dmabuf;
 	}

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

@@ -169,6 +169,16 @@ uint amdgpu_sdma_phase_quantum = 32;
 char *amdgpu_disable_cu;
 char *amdgpu_virtual_display;
 bool enforce_isolation;
+
+/* Specifies the default granularity for SVM, used in buffer
+ * migration and restoration of backing memory when handling
+ * recoverable page faults.
+ *
+ * The value is given as log(numPages(buffer)); for a 2 MiB
+ * buffer it computes to be 9
+ */
+uint amdgpu_svm_default_granularity = 9;
+
 /*
  * OverDrive(bit 14) disabled by default
  * GFX DCS(bit 19) disabled by default
@@ -320,6 +330,13 @@ module_param_named(pcie_gen2, amdgpu_pcie_gen2, int, 0444);
 MODULE_PARM_DESC(msi, "MSI support (1 = enable, 0 = disable, -1 = auto)");
 module_param_named(msi, amdgpu_msi, int, 0444);
 
+/**
+ * DOC: svm_default_granularity (uint)
+ * Used in buffer migration and handling of recoverable page faults
+ */
+MODULE_PARM_DESC(svm_default_granularity, "SVM's default granularity in log(2^Pages), default 9 = 2^9 = 2 MiB");
+module_param_named(svm_default_granularity, amdgpu_svm_default_granularity, uint, 0644);
+
 /**
  * DOC: lockup_timeout (string)
  * Set GPU scheduler timeout value in ms.

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

@@ -1397,14 +1397,23 @@ static ssize_t amdgpu_gfx_get_available_compute_partition(struct device *dev,
 static int amdgpu_gfx_run_cleaner_shader_job(struct amdgpu_ring *ring)
 {
 	struct amdgpu_device *adev = ring->adev;
-	long timeout = msecs_to_jiffies(1000);
-	struct dma_fence *f = NULL;
+	struct drm_gpu_scheduler *sched = &ring->sched;
+	struct drm_sched_entity entity;
+	struct dma_fence *f;
 	struct amdgpu_job *job;
 	struct amdgpu_ib *ib;
 	int i, r;
 
-	r = amdgpu_job_alloc_with_ib(adev, NULL, NULL,
-				     64, AMDGPU_IB_POOL_DIRECT,
+	/* Initialize the scheduler entity */
+	r = drm_sched_entity_init(&entity, DRM_SCHED_PRIORITY_NORMAL,
+				  &sched, 1, NULL);
+	if (r) {
+		dev_err(adev->dev, "Failed setting up GFX kernel entity.\n");
+		goto err;
+	}
+
+	r = amdgpu_job_alloc_with_ib(ring->adev, &entity, NULL,
+				     64, 0,
 				     &job);
 	if (r)
 		goto err;
@@ -1416,24 +1425,18 @@ static int amdgpu_gfx_run_cleaner_shader_job(struct amdgpu_ring *ring)
 		ib->ptr[i] = ring->funcs->nop;
 	ib->length_dw = ring->funcs->align_mask + 1;
 
-	r = amdgpu_job_submit_direct(job, ring, &f);
-	if (r)
-		goto err_free;
+	f = amdgpu_job_submit(job);
 
-	r = dma_fence_wait_timeout(f, false, timeout);
-	if (r == 0)
-		r = -ETIMEDOUT;
-	else if (r > 0)
-		r = 0;
+	r = dma_fence_wait(f, false);
+	if (r)
+		goto err;
 
-	amdgpu_ib_free(adev, ib, f);
 	dma_fence_put(f);
 
+	/* Clean up the scheduler entity */
+	drm_sched_entity_destroy(&entity);
 	return 0;
 
-err_free:
-	amdgpu_job_free(job);
-	amdgpu_ib_free(adev, ib, f);
 err:
 	return r;
 }

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

@@ -832,6 +832,7 @@ int amdgpu_mes_reset_hw_queue_mmio(struct amdgpu_device *adev, int queue_type,
 	struct mes_reset_queue_input queue_input;
 	int r;
 
+	queue_input.queue_type = queue_type;
 	queue_input.use_mmio = true;
 	queue_input.me_id = me_id;
 	queue_input.pipe_id = pipe_id;

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

@@ -94,7 +94,7 @@ static void amdgpu_pll_get_fb_ref_div(struct amdgpu_device *adev, unsigned int n
 		ref_div_max = min(128 / post_div, ref_div_max);
 
 	/* get matching reference and feedback divider */
-	*ref_div = min(max(DIV_ROUND_CLOSEST(den, post_div), 1u), ref_div_max);
+	*ref_div = clamp(DIV_ROUND_CLOSEST(den, post_div), 1u, ref_div_max);
 	*fb_div = DIV_ROUND_CLOSEST(nom * *ref_div * post_div, den);
 
 	/* limit fb divider to its maximum */

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

@@ -22,7 +22,6 @@
  * Authors: Andres Rodriguez <[email protected]>
  */
 
-#include <linux/fdtable.h>
 #include <linux/file.h>
 #include <linux/pid.h>