Merge tag 'amd-drm-next-6.2-2022-11-18' of https://gitlab.freedesktop.org/agd5f/linux into drm-next

amd-drm-next-6.2-2022-11-18:

amdgpu:
- SR-IOV fixes
- Clean up DC checks
- DCN 3.2.x fixes
- DCN 3.1.x fixes
- Don't enable degamma on asics which don't support it
- IP discovery fixes
- BACO fixes
- Fix vbios allocation handling when vkms is enabled
- Drop buggy tdr advanced mode GPU reset handling
- Fix the build when DCN is not set in kconfig
- MST DSC fixes
- Userptr fixes
- FRU and RAS EEPROM fixes
- VCN 4.x RAS support
- Aldrebaran CU occupancy reporting fix
- PSP ring cleanup

amdkfd:
- Memory limit fix
- Enable cooperative launch on gfx 10.3

amd-drm-next-6.2-2022-11-11:

amdgpu:
- SMU 13.x updates
- GPUVM TLB race fix
- DCN 3.1.4 updates
- DCN 3.2.x updates
- PSR fixes
- Kerneldoc fix
- Vega10 fan fix
- GPUVM locking fixes in error pathes
- BACO fix for Beige Goby
- EEPROM I2C address cleanup
- GFXOFF fix
- Fix DC memory leak in error pathes
- Flexible array updates
- Mtype fix for GPUVM PTEs
- Move Kconfig into amdgpu directory
- SR-IOV updates
- Fix possible memory leak in CS IOCTL error path

amdkfd:
- Fix possible memory overrun
- CRIU fixes

radeon:
- ACPI ref count fix
- HDA audio notifier support
- Move Kconfig into radeon directory

UAPI:
- Add new GEM_CREATE flags to help to transition more KFD functionality to the DRM UAPI.
  These are used internally in the driver to align location based memory coherency
  requirements from memory allocated in the KFD with how we manage GPUVM PTEs.  They
  are currently blocked in the GEM_CREATE IOCTL as we don't have a user right now.
  They are just used internally in the kernel driver for now for existing KFD memory
  allocations. So a change to the UAPI header, but no functional change in the UAPI.

From: Alex Deucher <alexander.deucher@amd.com>
Link: https://patchwork.freedesktop.org/patch/msgid/20221118170807.6505-1-alexander.deucher@amd.com
Signed-off-by: Dave Airlie <airlied@redhat.com>

1 files changed, 108 insertions, 107 deletions

diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_fru_eeprom.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_fru_eeprom.c
index e325150879df..2c38ac7bc643 100644
--- a/drivers/gpu/drm/amd/amdgpu/amdgpu_fru_eeprom.c
+++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_fru_eeprom.c
@@ -29,9 +29,10 @@
 #include "amdgpu_fru_eeprom.h"
 #include "amdgpu_eeprom.h"
 
-#define FRU_EEPROM_MADDR        0x60000
+#define FRU_EEPROM_MADDR_6      0x60000
+#define FRU_EEPROM_MADDR_8      0x80000
 
-static bool is_fru_eeprom_supported(struct amdgpu_device *adev)
+static bool is_fru_eeprom_supported(struct amdgpu_device *adev, u32 *fru_addr)
 {
 	/* Only server cards have the FRU EEPROM
 	 * TODO: See if we can figure this out dynamically instead of
@@ -45,6 +46,11 @@ static bool is_fru_eeprom_supported(struct amdgpu_device *adev)
 	if (amdgpu_sriov_vf(adev))
 		return false;
 
+	/* The default I2C EEPROM address of the FRU.
+	 */
+	if (fru_addr)
+		*fru_addr = FRU_EEPROM_MADDR_8;
+
 	/* VBIOS is of the format ###-DXXXYYYY-##. For SKU identification,
 	 * we can use just the "DXXX" portion. If there were more models, we
 	 * could convert the 3 characters to a hex integer and use a switch
@@ -57,21 +63,29 @@ static bool is_fru_eeprom_supported(struct amdgpu_device *adev)
 		if (strnstr(atom_ctx->vbios_version, "D161",
 			    sizeof(atom_ctx->vbios_version)) ||
 		    strnstr(atom_ctx->vbios_version, "D163",
-			    sizeof(atom_ctx->vbios_version)))
+			    sizeof(atom_ctx->vbios_version))) {
+			*fru_addr = FRU_EEPROM_MADDR_6;
 			return true;
-		else
+		} else {
 			return false;
+		}
 	case CHIP_ALDEBARAN:
-		/* All Aldebaran SKUs have the FRU */
+		/* All Aldebaran SKUs have an FRU */
+		if (!strnstr(atom_ctx->vbios_version, "D673",
+			     sizeof(atom_ctx->vbios_version)))
+			if (fru_addr)
+				*fru_addr = FRU_EEPROM_MADDR_6;
 		return true;
 	case CHIP_SIENNA_CICHLID:
 		if (strnstr(atom_ctx->vbios_version, "D603",
-		    sizeof(atom_ctx->vbios_version))) {
+			    sizeof(atom_ctx->vbios_version))) {
 			if (strnstr(atom_ctx->vbios_version, "D603GLXE",
-			    sizeof(atom_ctx->vbios_version)))
+				    sizeof(atom_ctx->vbios_version))) {
 				return false;
-			else
+			} else {
+				*fru_addr = FRU_EEPROM_MADDR_6;
 				return true;
+			}
 		} else {
 			return false;
 		}
@@ -80,41 +94,14 @@ static bool is_fru_eeprom_supported(struct amdgpu_device *adev)
 	}
 }
 
-static int amdgpu_fru_read_eeprom(struct amdgpu_device *adev, uint32_t addrptr,
-				  unsigned char *buf, size_t buf_size)
-{
-	int ret;
-	u8 size;
-
-	ret = amdgpu_eeprom_read(adev->pm.fru_eeprom_i2c_bus, addrptr, buf, 1);
-	if (ret < 1) {
-		DRM_WARN("FRU: Failed to get size field");
-		return ret;
-	}
-
-	/* The size returned by the i2c requires subtraction of 0xC0 since the
-	 * size apparently always reports as 0xC0+actual size.
-	 */
-	size = buf[0] & 0x3F;
-	size = min_t(size_t, size, buf_size);
-
-	ret = amdgpu_eeprom_read(adev->pm.fru_eeprom_i2c_bus, addrptr + 1,
-				 buf, size);
-	if (ret < 1) {
-		DRM_WARN("FRU: Failed to get data field");
-		return ret;
-	}
-
-	return size;
-}
-
 int amdgpu_fru_get_product_info(struct amdgpu_device *adev)
 {
-	unsigned char buf[AMDGPU_PRODUCT_NAME_LEN];
-	u32 addrptr;
+	unsigned char buf[8], *pia;
+	u32 addr, fru_addr;
 	int size, len;
+	u8 csum;
 
-	if (!is_fru_eeprom_supported(adev))
+	if (!is_fru_eeprom_supported(adev, &fru_addr))
 		return 0;
 
 	/* If algo exists, it means that the i2c_adapter's initialized */
@@ -123,88 +110,102 @@ int amdgpu_fru_get_product_info(struct amdgpu_device *adev)
 		return -ENODEV;
 	}
 
-	/* There's a lot of repetition here. This is due to the FRU having
-	 * variable-length fields. To get the information, we have to find the
-	 * size of each field, and then keep reading along and reading along
-	 * until we get all of the data that we want. We use addrptr to track
-	 * the address as we go
-	 */
-
-	/* The first fields are all of size 1-byte, from 0-7 are offsets that
-	 * contain information that isn't useful to us.
-	 * Bytes 8-a are all 1-byte and refer to the size of the entire struct,
-	 * and the language field, so just start from 0xb, manufacturer size
-	 */
-	addrptr = FRU_EEPROM_MADDR + 0xb;
-	size = amdgpu_fru_read_eeprom(adev, addrptr, buf, sizeof(buf));
-	if (size < 1) {
-		DRM_ERROR("Failed to read FRU Manufacturer, ret:%d", size);
-		return -EINVAL;
+	/* Read the IPMI Common header */
+	len = amdgpu_eeprom_read(adev->pm.fru_eeprom_i2c_bus, fru_addr, buf,
+				 sizeof(buf));
+	if (len != 8) {
+		DRM_ERROR("Couldn't read the IPMI Common Header: %d", len);
+		return len < 0 ? len : -EIO;
 	}
 
-	/* Increment the addrptr by the size of the field, and 1 due to the
-	 * size field being 1 byte. This pattern continues below.
-	 */
-	addrptr += size + 1;
-	size = amdgpu_fru_read_eeprom(adev, addrptr, buf, sizeof(buf));
-	if (size < 1) {
-		DRM_ERROR("Failed to read FRU product name, ret:%d", size);
-		return -EINVAL;
+	if (buf[0] != 1) {
+		DRM_ERROR("Bad IPMI Common Header version: 0x%02x", buf[0]);
+		return -EIO;
 	}
 
-	len = size;
-	if (len >= AMDGPU_PRODUCT_NAME_LEN) {
-		DRM_WARN("FRU Product Name is larger than %d characters. This is likely a mistake",
-				AMDGPU_PRODUCT_NAME_LEN);
-		len = AMDGPU_PRODUCT_NAME_LEN - 1;
-	}
-	memcpy(adev->product_name, buf, len);
-	adev->product_name[len] = '\0';
-
-	addrptr += size + 1;
-	size = amdgpu_fru_read_eeprom(adev, addrptr, buf, sizeof(buf));
-	if (size < 1) {
-		DRM_ERROR("Failed to read FRU product number, ret:%d", size);
-		return -EINVAL;
+	for (csum = 0; len > 0; len--)
+		csum += buf[len - 1];
+	if (csum) {
+		DRM_ERROR("Bad IPMI Common Header checksum: 0x%02x", csum);
+		return -EIO;
 	}
 
-	len = size;
-	/* Product number should only be 16 characters. Any more,
-	 * and something could be wrong. Cap it at 16 to be safe
-	 */
-	if (len >= sizeof(adev->product_number)) {
-		DRM_WARN("FRU Product Number is larger than 16 characters. This is likely a mistake");
-		len = sizeof(adev->product_number) - 1;
-	}
-	memcpy(adev->product_number, buf, len);
-	adev->product_number[len] = '\0';
+	/* Get the offset to the Product Info Area (PIA). */
+	addr = buf[4] * 8;
+	if (!addr)
+		return 0;
 
-	addrptr += size + 1;
-	size = amdgpu_fru_read_eeprom(adev, addrptr, buf, sizeof(buf));
+	/* Get the absolute address to the PIA. */
+	addr += fru_addr;
 
-	if (size < 1) {
-		DRM_ERROR("Failed to read FRU product version, ret:%d", size);
-		return -EINVAL;
+	/* Read the header of the PIA. */
+	len = amdgpu_eeprom_read(adev->pm.fru_eeprom_i2c_bus, addr, buf, 3);
+	if (len != 3) {
+		DRM_ERROR("Couldn't read the Product Info Area header: %d", len);
+		return len < 0 ? len : -EIO;
 	}
 
-	addrptr += size + 1;
-	size = amdgpu_fru_read_eeprom(adev, addrptr, buf, sizeof(buf));
+	if (buf[0] != 1) {
+		DRM_ERROR("Bad IPMI Product Info Area version: 0x%02x", buf[0]);
+		return -EIO;
+	}
 
-	if (size < 1) {
-		DRM_ERROR("Failed to read FRU serial number, ret:%d", size);
-		return -EINVAL;
+	size = buf[1] * 8;
+	pia = kzalloc(size, GFP_KERNEL);
+	if (!pia)
+		return -ENOMEM;
+
+	/* Read the whole PIA. */
+	len = amdgpu_eeprom_read(adev->pm.fru_eeprom_i2c_bus, addr, pia, size);
+	if (len != size) {
+		kfree(pia);
+		DRM_ERROR("Couldn't read the Product Info Area: %d", len);
+		return len < 0 ? len : -EIO;
 	}
 
-	len = size;
-	/* Serial number should only be 16 characters. Any more,
-	 * and something could be wrong. Cap it at 16 to be safe
-	 */
-	if (len >= sizeof(adev->serial)) {
-		DRM_WARN("FRU Serial Number is larger than 16 characters. This is likely a mistake");
-		len = sizeof(adev->serial) - 1;
+	for (csum = 0; size > 0; size--)
+		csum += pia[size - 1];
+	if (csum) {
+		DRM_ERROR("Bad Product Info Area checksum: 0x%02x", csum);
+		return -EIO;
 	}
-	memcpy(adev->serial, buf, len);
-	adev->serial[len] = '\0';
 
+	/* Now extract useful information from the PIA.
+	 *
+	 * Skip the Manufacturer Name at [3] and go directly to
+	 * the Product Name field.
+	 */
+	addr = 3 + 1 + (pia[3] & 0x3F);
+	if (addr + 1 >= len)
+		goto Out;
+	memcpy(adev->product_name, pia + addr + 1,
+	       min_t(size_t,
+		     sizeof(adev->product_name),
+		     pia[addr] & 0x3F));
+	adev->product_name[sizeof(adev->product_name) - 1] = '\0';
+
+	/* Go to the Product Part/Model Number field. */
+	addr += 1 + (pia[addr] & 0x3F);
+	if (addr + 1 >= len)
+		goto Out;
+	memcpy(adev->product_number, pia + addr + 1,
+	       min_t(size_t,
+		     sizeof(adev->product_number),
+		     pia[addr] & 0x3F));
+	adev->product_number[sizeof(adev->product_number) - 1] = '\0';
+
+	/* Go to the Product Version field. */
+	addr += 1 + (pia[addr] & 0x3F);
+
+	/* Go to the Product Serial Number field. */
+	addr += 1 + (pia[addr] & 0x3F);
+	if (addr + 1 >= len)
+		goto Out;
+	memcpy(adev->serial, pia + addr + 1, min_t(size_t,
+						   sizeof(adev->serial),
+						   pia[addr] & 0x3F));
+	adev->serial[sizeof(adev->serial) - 1] = '\0';
+Out:
+	kfree(pia);
 	return 0;
 }