On 5/22/24 6:37 AM, Jayesh Choudhary wrote:
Enable Quality of Service (QoS) blocks for Display SubSystem (DSS), by
servicing the DSS - DDR traffic from the Real-Time (RT) queue. This is
done by setting the DSS DMA orderID to greater than 7.
ATYPE 3 is selected so that the traffic takes non-coherent path and does
So this is the part I'm not sure about, this change is not just a performance
adjustment, it can and does break things. The ORDERID setting is fine with me,
but changing ATYPE should be done in a separate patch so it can be debated on
its own.
not have a conflict with coherent traffic from C7x (deep-learning
applications).
Before setting up the QoS, the ORDERID needs to be mapped to VBUSM sources
using setup_navss_nb() function call that sets the threadmap for NBSS
registers. (Section 10.2.10.1.2 "NB Parameters" in TRM[0])
Section 3.3.2 "Quality of Service (QoS)" in the TRM[0] provide more
details.
[0]: https://www.ti.com/lit/zip/spruil1
Signed-off-by: Jayesh Choudhary <j-choudh...@ti.com>
---
arch/arm/mach-k3/j721e/j721e_init.c | 28 +++++
arch/arm/mach-k3/r5/j721e/Makefile | 1 +
arch/arm/mach-k3/r5/j721e/j721e_qos.h | 96 +++++++++++++++
arch/arm/mach-k3/r5/j721e/j721e_qos_uboot.c | 125 ++++++++++++++++++++
4 files changed, 250 insertions(+)
create mode 100644 arch/arm/mach-k3/r5/j721e/j721e_qos.h
create mode 100644 arch/arm/mach-k3/r5/j721e/j721e_qos_uboot.c
diff --git a/arch/arm/mach-k3/j721e/j721e_init.c
b/arch/arm/mach-k3/j721e/j721e_init.c
index c2024f2500..e9ed8cb267 100644
--- a/arch/arm/mach-k3/j721e/j721e_init.c
+++ b/arch/arm/mach-k3/j721e/j721e_init.c
@@ -23,6 +23,22 @@
#include "../sysfw-loader.h"
#include "../common.h"
+/* NAVSS North Bridge (NB) registers */
+#define NAVSS0_NBSS_NB0_CFG_MMRS 0x03802000
+#define NAVSS0_NBSS_NB1_CFG_MMRS 0x03803000
+#define NAVSS0_NBSS_NB0_CFG_NB_THREADMAP (NAVSS0_NBSS_NB0_CFG_MMRS +
0x10)
+#define NAVSS0_NBSS_NB1_CFG_NB_THREADMAP (NAVSS0_NBSS_NB1_CFG_MMRS +
0x10)
+/*
+ * Thread Map for North Bridge Configuration
+ * Each bit is for each VBUSM source.
+ * Bit[0] maps orderID 0-7 to VBUSM.C thread number
+ * Bit[1] maps orderID 8-15 to VBUSM.C thread number
+ * When bit has value 0: VBUSM.C thread 0 (non-real time traffic)
+ * When bit has value 1: VBUSM.C thread 2 (real time traffic)
+ */
+#define NB_THREADMAP_BIT0 BIT(0)
+#define NB_THREADMAP_BIT1 BIT(1)
+
#ifdef CONFIG_K3_LOAD_SYSFW
struct fwl_data cbass_hc_cfg0_fwls[] = {
#if defined(CONFIG_TARGET_J721E_R5_EVM)
@@ -124,6 +140,13 @@ void k3_mmc_restart_clock(void)
}
#endif
+/* Setup North Bridge registers to map ORDERID 8-15 to RT traffic */
+static void setup_navss_nb(void)
+{
+ writel(NB_THREADMAP_BIT1, (uintptr_t)NAVSS0_NBSS_NB0_CFG_NB_THREADMAP);
+ writel(NB_THREADMAP_BIT1, (uintptr_t)NAVSS0_NBSS_NB1_CFG_NB_THREADMAP);
+}
+
/*
* This uninitialized global variable would normal end up in the .bss section,
* but the .bss is cleared between writing and reading this variable, so move
@@ -288,6 +311,11 @@ void board_init_f(ulong dummy)
panic("DRAM init failed: %d\n", ret);
#endif
spl_enable_cache();
+
+ if (IS_ENABLED(CONFIG_CPU_V7R))
+ setup_navss_nb();
+
+ setup_qos();
}
u32 spl_mmc_boot_mode(struct mmc *mmc, const u32 boot_device)
diff --git a/arch/arm/mach-k3/r5/j721e/Makefile
b/arch/arm/mach-k3/r5/j721e/Makefile
index 78325db402..07bfb0dd93 100644
--- a/arch/arm/mach-k3/r5/j721e/Makefile
+++ b/arch/arm/mach-k3/r5/j721e/Makefile
@@ -3,3 +3,4 @@
# Copyright (C) 2020 Texas Instruments Incorporated - https://www.ti.com/
obj-y += clk-data.o
obj-y += dev-data.o
+obj-y += j721e_qos_uboot.o
diff --git a/arch/arm/mach-k3/r5/j721e/j721e_qos.h
b/arch/arm/mach-k3/r5/j721e/j721e_qos.h
new file mode 100644
index 0000000000..9ec0b7c630
--- /dev/null
+++ b/arch/arm/mach-k3/r5/j721e/j721e_qos.h
@@ -0,0 +1,96 @@
+/* SPDX-License-Identifier: GPL-2.0+ */
+/*
+ * Keystone3 Quality of service endpoint definitions
+ * Auto generated by K3 Resource Partitioning Tool
+ *
+ * Copyright (C) 2024 Texas Instruments Incorporated - https://www.ti.com/
+ */
+
+#define PULSAR_SL_MCU_0_MEMBDG_RMST0 0x45D10000
+#define PULSAR_SL_MCU_0_MEMBDG_WMST0 0x45D10400
+#define PULSAR_SL_MCU_0_CPU0_PMST 0x45D10800
+#define PULSAR_SL_MCU_0_MEMBDG_RMST1 0x45D11000
+#define PULSAR_SL_MCU_0_MEMBDG_WMST1 0x45D11400
+#define PULSAR_SL_MCU_0_CPU1_PMST 0x45D11800
+#define SA2_UL_MCU_0_CTXCACH_EXT_DMA 0x45D13000
+#define ICSS_G_MAIN_0_PR1_EXT_VBUSM 0x45D80000
+#define ICSS_G_MAIN_1_PR1_EXT_VBUSM 0x45D80400
+#define K3_C66_COREPAC_MAIN_0_C66_MDMA 0x45D81000
+#define K3_C66_COREPAC_MAIN_1_C66_MDMA 0x45D81400
+#define EMMCSD4SS_MAIN_0_EMMCSDSS_RD 0x45D82000
+#define EMMCSD4SS_MAIN_0_EMMCSDSS_WR 0x45D82400
+#define EMMCSD4SS_MAIN_1_EMMCSDSS_RD 0x45D82800
+#define EMMCSD4SS_MAIN_1_EMMCSDSS_WR 0x45D82C00
+#define PULSAR_SL_MAIN_0_MEMBDG_RMST0 0x45D84000
+#define PULSAR_SL_MAIN_0_MEMBDG_RMST1 0x45D84400
+#define PULSAR_SL_MAIN_0_MEMBDG_WMST0 0x45D84800
+#define PULSAR_SL_MAIN_0_MEMBDG_WMST1 0x45D84C00
+#define PULSAR_SL_MAIN_1_MEMBDG_RMST0 0x45D85000
+#define PULSAR_SL_MAIN_1_MEMBDG_RMST1 0x45D85400
+#define PULSAR_SL_MAIN_1_MEMBDG_WMST0 0x45D85800
+#define PULSAR_SL_MAIN_1_MEMBDG_WMST1 0x45D85C00
+#define COMPUTE_CLUSTER_J7ES_TB_VDC_MAIN_0_GIC_MEM_RD_VBUSM 0x45D86000
+#define COMPUTE_CLUSTER_J7ES_TB_VDC_MAIN_0_GIC_MEM_WR_VBUSM 0x45D86400
+#define K3_C66_COREPAC_MAIN_0_C66_CFG 0x45D87000
+#define K3_C66_COREPAC_MAIN_1_C66_CFG 0x45D87400
+#define SA2_UL_MAIN_0_CTXCACH_EXT_DMA 0x45D88800
+#define PULSAR_SL_MAIN_0_PBDG_RMST0 0x45D89800
+#define PULSAR_SL_MAIN_0_PBDG_RMST1 0x45D89C00
+#define PULSAR_SL_MAIN_0_PBDG_WMST0 0x45D8A000
+#define PULSAR_SL_MAIN_0_PBDG_WMST1 0x45D8A400
+#define PULSAR_SL_MAIN_1_PBDG_RMST0 0x45D8A800
+#define PULSAR_SL_MAIN_1_PBDG_RMST1 0x45D8AC00
+#define PULSAR_SL_MAIN_1_PBDG_WMST0 0x45D8B000
+#define PULSAR_SL_MAIN_1_PBDG_WMST1 0x45D8B400
+#define VPFE_MAIN_0_VBUSM_DMA 0x45D8C000
+#define VPE_MAIN_0_VPDMA_MST0 0x45D8C400
+#define VPE_MAIN_0_VPDMA_MST1 0x45D8C800
+#define PCIE_G4X2_MAIN_0_PCIE_MST_RD_HP 0x45D90000
+#define PCIE_G4X2_MAIN_0_PCIE_MST_RD_LP 0x45D90400
+#define PCIE_G4X2_MAIN_0_PCIE_MST_WR_HP 0x45D90800
+#define PCIE_G4X2_MAIN_0_PCIE_MST_WR_LP 0x45D90C00
+#define PCIE_G4X2_MAIN_1_PCIE_MST_RD_HP 0x45D91000
+#define PCIE_G4X2_MAIN_1_PCIE_MST_RD_LP 0x45D91400
+#define PCIE_G4X2_MAIN_1_PCIE_MST_WR_HP 0x45D91800
+#define PCIE_G4X2_MAIN_1_PCIE_MST_WR_LP 0x45D91C00
+#define PCIE_G4X2_MAIN_2_PCIE_MST_RD_HP 0x45D92000
+#define PCIE_G4X2_MAIN_2_PCIE_MST_RD_LP 0x45D92400
+#define PCIE_G4X2_MAIN_2_PCIE_MST_WR_HP 0x45D92800
+#define PCIE_G4X2_MAIN_2_PCIE_MST_WR_LP 0x45D92C00
+#define PCIE_G4X2_MAIN_3_PCIE_MST_RD_HP 0x45D93000
+#define PCIE_G4X2_MAIN_3_PCIE_MST_RD_LP 0x45D93400
+#define PCIE_G4X2_MAIN_3_PCIE_MST_WR_HP 0x45D93800
+#define PCIE_G4X2_MAIN_3_PCIE_MST_WR_LP 0x45D93C00
+#define USB3P0SS_16FFC_MAIN_0_MSTR0 0x45D98000
+#define USB3P0SS_16FFC_MAIN_0_MSTW0 0x45D98400
+#define USB3P0SS_16FFC_MAIN_1_MSTR0 0x45D98800
+#define USB3P0SS_16FFC_MAIN_1_MSTW0 0x45D98C00
+#define USB3P0SS_16FFC_MAIN_2_MSTR0 0x45D99000
+#define USB3P0SS_16FFC_MAIN_2_MSTW0 0x45D99400
+#define MLBSS2P0_MAIN_0_MLBSS_DMA_VBUSP 0x45D99C00
+#define EMMC8SS_16FFC_MAIN_0_EMMCSS_RD 0x45D9A000
+#define EMMC8SS_16FFC_MAIN_0_EMMCSS_WR 0x45D9A400
+#define UFSHCI2P1SS_16FFC_MAIN_0_UFSHCI_VBM_MST_RD 0x45D9B000
+#define UFSHCI2P1SS_16FFC_MAIN_0_UFSHCI_VBM_MST_WR 0x45D9B400
+#define UFSHCI2P1SS_16FFC_MAIN_1_UFSHCI_VBM_MST_RD 0x45D9B800
+#define UFSHCI2P1SS_16FFC_MAIN_1_UFSHCI_VBM_MST_WR 0x45D9BC00
+#define DEBUGSS_K3_WRAP_CV0_MAIN_0_VBUSMR 0x45DA0000
+#define DEBUGSS_K3_WRAP_CV0_MAIN_0_VBUSMW 0x45DA0400
+#define PULSAR_SL_MAIN_0_CPU0_PMST 0x45DA4000
+#define PULSAR_SL_MAIN_0_CPU1_PMST 0x45DA4400
+#define PULSAR_SL_MAIN_1_CPU0_PMST 0x45DA4800
+#define PULSAR_SL_MAIN_1_CPU1_PMST 0x45DA4C00
+#define DMPAC_TOP_MAIN_0_DATA_MST 0x45DC0000
+#define K3_D5520MP2_MAIN_0_M_VBUSM_R 0x45DC0400
+#define K3_D5520MP2_MAIN_0_M_VBUSM_W 0x45DC0800
+#define K3_VXE384MP2_MAIN_0_M_VBUSM_R 0x45DC0C00
+#define K3_VXE384MP2_MAIN_0_M_VBUSM_W 0x45DC1000
+#define VPAC_TOP_MAIN_0_DATA_MST_0 0x45DC1400
+#define VPAC_TOP_MAIN_0_DATA_MST_1 0x45DC1800
+#define VPAC_TOP_MAIN_0_LDC0_M_MST 0x45DC1C00
+#define K3_DSS_MAIN_0_DSS_INST0_VBUSM_DMA 0x45DC2000
+#define K3_DSS_MAIN_0_DSS_INST0_VBUSM_FBDC 0x45DC2400
+#define J7_LASCAR_GPU_WRAP_MAIN_0_M0_VBUSM_R_ASYNC 0x45DC5000
+#define J7_LASCAR_GPU_WRAP_MAIN_0_M0_VBUSM_W_ASYNC 0x45DC5800
+#define J7_LASCAR_GPU_WRAP_MAIN_0_M1_VBUSM_R_ASYNC 0x45DC6000
+#define J7_LASCAR_GPU_WRAP_MAIN_0_M1_VBUSM_W_ASYNC 0x45DC6800
diff --git a/arch/arm/mach-k3/r5/j721e/j721e_qos_uboot.c
b/arch/arm/mach-k3/r5/j721e/j721e_qos_uboot.c
new file mode 100644
index 0000000000..c829057200
--- /dev/null
+++ b/arch/arm/mach-k3/r5/j721e/j721e_qos_uboot.c
@@ -0,0 +1,125 @@
+// SPDX-License-Identifier: GPL-2.0+
+/*
+ * j721e Quality of Service (QoS) Configuration Data
+ * Auto generated from K3 Resource Partitioning tool
+ */
+
+#include <asm/arch/k3-qos.h>
+#include "j721e_qos.h"
+
+struct k3_qos_data qos_data[] = {
+ /* DSS_PIPE_VID1 - 2 endpoints, 2 channels */
+ {
+ .reg = K3_DSS_MAIN_0_DSS_INST0_VBUSM_DMA + 0x100 + 0x4 * 0,
0x100 being the instance, and 0x4 * x being the specific device register right?
That could be encoded in a macro (the "Resource Partitioning tool" would
be updated to generate that too).
Andrew
+ .val = ATYPE_3 | ORDERID_15,
+ },
+ {
+ .reg = K3_DSS_MAIN_0_DSS_INST0_VBUSM_DMA + 0x100 + 0x4 * 1,
+ .val = ATYPE_3 | ORDERID_15,
+ },
+ {
+ .reg = K3_DSS_MAIN_0_DSS_INST0_VBUSM_FBDC + 0x100 + 0x4 * 0,
+ .val = ATYPE_3 | ORDERID_15,
+ },
+ {
+ .reg = K3_DSS_MAIN_0_DSS_INST0_VBUSM_FBDC + 0x100 + 0x4 * 1,
+ .val = ATYPE_3 | ORDERID_15,
+ },
+
+ /* DSS_PIPE_VIDL1 - 2 endpoints, 2 channels */
+ {
+ .reg = K3_DSS_MAIN_0_DSS_INST0_VBUSM_DMA + 0x100 + 0x4 * 2,
+ .val = ATYPE_3 | ORDERID_15,
+ },
+ {
+ .reg = K3_DSS_MAIN_0_DSS_INST0_VBUSM_DMA + 0x100 + 0x4 * 3,
+ .val = ATYPE_3 | ORDERID_15,
+ },
+ {
+ .reg = K3_DSS_MAIN_0_DSS_INST0_VBUSM_FBDC + 0x100 + 0x4 * 2,
+ .val = ATYPE_3 | ORDERID_15,
+ },
+ {
+ .reg = K3_DSS_MAIN_0_DSS_INST0_VBUSM_FBDC + 0x100 + 0x4 * 3,
+ .val = ATYPE_3 | ORDERID_15,
+ },
+
+ /* DSS_PIPE_VID2 - 2 endpoints, 2 channels */
+ {
+ .reg = K3_DSS_MAIN_0_DSS_INST0_VBUSM_DMA + 0x100 + 0x4 * 4,
+ .val = ATYPE_3 | ORDERID_15,
+ },
+ {
+ .reg = K3_DSS_MAIN_0_DSS_INST0_VBUSM_DMA + 0x100 + 0x4 * 5,
+ .val = ATYPE_3 | ORDERID_15,
+ },
+ {
+ .reg = K3_DSS_MAIN_0_DSS_INST0_VBUSM_FBDC + 0x100 + 0x4 * 4,
+ .val = ATYPE_3 | ORDERID_15,
+ },
+ {
+ .reg = K3_DSS_MAIN_0_DSS_INST0_VBUSM_FBDC + 0x100 + 0x4 * 5,
+ .val = ATYPE_3 | ORDERID_15,
+ },
+
+ /* DSS_PIPE_VIDL2 - 2 endpoints, 2 channels */
+ {
+ .reg = K3_DSS_MAIN_0_DSS_INST0_VBUSM_DMA + 0x100 + 0x4 * 6,
+ .val = ATYPE_3 | ORDERID_15,
+ },
+ {
+ .reg = K3_DSS_MAIN_0_DSS_INST0_VBUSM_DMA + 0x100 + 0x4 * 7,
+ .val = ATYPE_3 | ORDERID_15,
+ },
+ {
+ .reg = K3_DSS_MAIN_0_DSS_INST0_VBUSM_FBDC + 0x100 + 0x4 * 6,
+ .val = ATYPE_3 | ORDERID_15,
+ },
+ {
+ .reg = K3_DSS_MAIN_0_DSS_INST0_VBUSM_FBDC + 0x100 + 0x4 * 7,
+ .val = ATYPE_3 | ORDERID_15,
+ },
+
+ /* Following registers set 1:1 mapping for orderID MAP1/MAP2
+ * remap registers. orderID x is remapped to orderID x again
+ * This is to ensure orderID from MAP register is unchanged
+ */
+
+ /* K3_DSS_MAIN_0_DSS_INST0_VBUSM_DMA - 2 groups */
+ {
+ .reg = K3_DSS_MAIN_0_DSS_INST0_VBUSM_DMA + 0,
+ .val = 0x76543210,
+ },
+ {
+ .reg = K3_DSS_MAIN_0_DSS_INST0_VBUSM_DMA + 4,
+ .val = 0xfedcba98,
+ },
+ {
+ .reg = K3_DSS_MAIN_0_DSS_INST0_VBUSM_DMA + 8,
+ .val = 0x76543210,
+ },
+ {
+ .reg = K3_DSS_MAIN_0_DSS_INST0_VBUSM_DMA + 12,
+ .val = 0xfedcba98,
+ },
+
+ /* K3_DSS_MAIN_0_DSS_INST0_VBUSM_FBDC - 2 groups */
+ {
+ .reg = K3_DSS_MAIN_0_DSS_INST0_VBUSM_FBDC + 0,
+ .val = 0x76543210,
+ },
+ {
+ .reg = K3_DSS_MAIN_0_DSS_INST0_VBUSM_FBDC + 4,
+ .val = 0xfedcba98,
+ },
+ {
+ .reg = K3_DSS_MAIN_0_DSS_INST0_VBUSM_FBDC + 8,
+ .val = 0x76543210,
+ },
+ {
+ .reg = K3_DSS_MAIN_0_DSS_INST0_VBUSM_FBDC + 12,
+ .val = 0xfedcba98,
+ },
+};
+
+u32 qos_count = ARRAY_SIZE(qos_data);
