[参考译文] AM62A7-Q1：偶尔出现 SBL 身份验证失败

尊敬的 Kevin：

Bootloader_verifyMulticoreImage 函数中存在多个故障点。 可能是映像本身在 DDR 中读取不正确的情况、因此即使在向 TIFS 发出任何映像身份验证请求之前、也可能会错误地读取证书长度、从而导致失败。

您能否要求客户在身份验证失败时检查映像长度(imageLen)和证书长度(certLen)的值是多少。

此致、

Prashant

您好、Prashant  

我检查一下 imageLen 和 certLen、它们是有效的。  

Bootloader_socAuthImage 返回值函数的错误为0xffffffff

出现"Debug (调试)"日志  

[sbl_emmc_linux_stage1] step 1
[sbl_emmc_linux_stage1] step 2
[sbl_emmc_linux_stage1] step 3
[sbl emmc linux stage1] Enter App_loadImages Functions
[sbl emmc linux stage1] App_loadImages step 1
[Bootloader] Enter Function of Bootloader_parseMultiCoreAppImage...
[Bootloader:Bootloader_parseMultiCoreAppImage] HS Device : VerifyMulticoreImage
[Bootloader] Enter Function of Bootloader_verifyMulticoreImage...
[Bootloader:Bootloader_verifyMulticoreImage] The bootMedia is 0xb0070003
[Bootloader:Bootloader_verifyMulticoreImage] The bootMedia is Flash or EMMC
[Bootloader:Bootloader_verifyMulticoreImage] certLen = 0x676
[Bootloader:Bootloader_verifyMulticoreImage] imageLen = 0xe754
[Bootloader:Bootloader_verifyMulticoreImage] authStatus = 0xffffffff
ERROR: Bootloader_verifyMulticoreImage:660: Failed to authenticate Image
[sbl emmc linux stage1] App_loadImages step 2
[sbl emmc linux stage1] App_loadImages step 3
[sbl emmc linux stage1] App_loadImages step 6
[sbl_emmc_linux_stage1] Some tests have failed!!

您好、Prashant

我在 Sciclient_procBootAuthAndStart 函数中找到问题位置

我在后面添加了一些调试信息  

int32_t Sciclient_procBootAuthAndStart(
            const struct tisci_msg_proc_auth_boot_req * authBootCfg,
            uint32_t timeout)
{
    int32_t retVal = CSL_PASS;

    struct tisci_msg_proc_auth_boot_resp response = {0};
    DebugP_log("[DM-sciclient] Enter Function of Sciclient_procBootAuthAndStart... \r\n");
    Sciclient_ReqPrm_t reqParam = {0};
    reqParam.messageType    = (uint16_t) TISCI_MSG_PROC_AUTH_BOOT;
    reqParam.flags          = (uint32_t) TISCI_MSG_FLAG_AOP;
    reqParam.pReqPayload    = (const uint8_t *) authBootCfg;
    reqParam.reqPayloadSize = (uint32_t) sizeof (struct tisci_msg_proc_auth_boot_req);
    reqParam.timeout        = (uint32_t) timeout;

    Sciclient_RespPrm_t respParam = {0};
    respParam.flags           = (uint32_t) 0;   /* Populated by the API */
    respParam.pRespPayload    = (uint8_t *) &response;
    respParam.respPayloadSize = (uint32_t) sizeof (response);

    retVal = Sciclient_service(&reqParam, &respParam);
    DebugP_log("[Bootloader:Sciclient_procBootAuthAndStart]  retVal = 0x%x,respParam.flags = 0x%x\r\n",retVal,respParam.flags);
    if((retVal != CSL_PASS) ||
        ((respParam.flags & TISCI_MSG_FLAG_ACK) != TISCI_MSG_FLAG_ACK))
    {
        retVal = CSL_EFAIL;
    }
    return retVal;
}

和调试输出信息

[sbl_emmc_linux_stage1] step 1
[sbl_emmc_linux_stage1] step 2
[sbl_emmc_linux_stage1] step 3
[sbl emmc linux stage1] Enter App_loadImages Functions
[sbl emmc linux stage1] App_loadImages step 1
[Bootloader] Enter Function of Bootloader_parseMultiCoreAppImage...
[Bootloader:Bootloader_parseMultiCoreAppImage] HS Device : VerifyMulticoreImage
[Bootloader] Enter Function of Bootloader_verifyMulticoreImage...
[Bootloader:Bootloader_verifyMulticoreImage] The bootMedia is 0xb0070003
[Bootloader:Bootloader_verifyMulticoreImage] The bootMedia is Flash or EMMC
[Bootloader:Bootloader_verifyMulticoreImage] certLen = 0x676
[Bootloader:Bootloader_verifyMulticoreImage] imageLen = 0xe754
[Bootloader:Bootloader_verifyMulticoreImage] certLoadAddr = 0x84000000
[Bootloader] Enter Function of Bootloader_socAuthImage...
[DM-sciclient] Enter Function of Sciclient_procBootAuthAndStart...
[DM-sciclient] Enter Function of Sciclient_service...
[DM-sciclient] Running the function of Sciclient_serviceGetThreadIds and ret = 0x0...
[DM-sciclient] Running the function of Sciclient_servicePrepareHeader and ret = 0x0...
[DM-sciclient] msgType = 0xc120...
[Bootloader:Sciclient_procBootAuthAndStart]  retVal = 0x0,respParam.flags = 0x0
[Bootloader:Bootloader_verifyMulticoreImage] authStatus = 0xffffffff
ERROR: Bootloader_verifyMulticoreImage:661: Failed to authenticate Image
[sbl emmc linux stage1] App_loadImages step 2
[sbl emmc linux stage1] App_loadImages step 3
[sbl emmc linux stage1] App_loadImages step 6
[sbl_emmc_linux_stage1] Some tests have failed!!

您好、Tom、

感谢您提供详细的调试。

在进行进一步讨论前、我想作一些澄清。

该问题是在对 sbl_uart_uniflash_stage2 appimage 进行身份验证时首次报告的。 但是、最新的日志显示 sbl_eMMC_linux_stage1可能无法验证 MCU appimage。 那么、您是否在这两种情况下都看到了身份验证失败？

此外、只是为了确认身份验证失败是随机的、而不是每次都发生？

此外、您使用的是定制电路板还是 TI AM62A EVM？

此致、

Prashant

您好、Prashant

是、在 sbl_uart_uniflash_stage1和 sbl_eMMC_linux_stage1中发现的问题。

我测试了 EVM、未发生此问题。

BR

Tom

Unknown 说：

我测试了 EVM、未发生此问题。

您的意思是此问题仅发生在您的定制电路板上吗？

另外、您没有说明身份验证失败是随机的还是在每次 POR 上发生的？ 如果是随机的、您能告诉随机故障的百分比吗？

是的、在定制电路板中会发生这种情况。 这是随机故障。未添加调试信息时、发生的概率大约为50%、并且在添加调试信息后、发生的概率会下降到大约10%

i debuging、因为需要 sciclient、但它不会响应 ACK、因此它失败。请参阅上面的调试信息

您好、Tom、

您能否按此处所述启用并共享 TIFS 日志

https://software-dl.ti.com/mcu-plus-sdk/esd/AM62AX/09_02_00_38/exports/docs/api_guide_am62ax/TOOLS_SYSFW.html#SYSFW_TRACE_ENABLE

您好、Prashant

我使用该指南、它会报告错误、后面显示错误信息

 28/05/2024   16:22.14   /drives/c/ti/mcu_plus_sdk_am62ax_09_01_00_39   master  gmake -s -C tools/sysfw/boardcfg SOC=am62ax
"Build Board Configuration Obj Files"
"Create unsigned Board Configuration binary"
perl: warning: Setting locale failed.
perl: warning: Please check that your locale settings:
        LC_ALL = "zh_CN.UTF-8",
        LC_CTYPE = "zh_CN.UTF-8",
        LANG = "zh_CN.UTF-8"
    are supported and installed on your system.
perl: warning: Falling back to the system default locale ("Chinese (Simplified)_China.936").
Load address not computed for file that does not have a segment table at XML_TI_OFD.pm line 1441.
Use of uninitialized value in multiplication (*) at script/mkhex4bin.pl line 295.
"Sign Board Configuration with GP key"
"  (note: GP board configuration not ACTUALLY signed with anything)"
"Generate include-able header file from signed Board Configuration"
"Build Board Configuration Obj Files"
"Create unsigned Board Configuration binary"
perl: warning: Setting locale failed.
perl: warning: Please check that your locale settings:
        LC_ALL = "zh_CN.UTF-8",
        LC_CTYPE = "zh_CN.UTF-8",
        LANG = "zh_CN.UTF-8"
    are supported and installed on your system.
perl: warning: Falling back to the system default locale ("Chinese (Simplified)_China.936").
Load address not computed for file that does not have a segment table at XML_TI_OFD.pm line 1441.
Use of uninitialized value in multiplication (*) at script/mkhex4bin.pl line 295.
"Sign Board Configuration with GP key"
"  (note: GP board configuration not ACTUALLY signed with anything)"
"Generate include-able header file from signed Board Configuration"
"Build Board Configuration Obj Files"
"Create unsigned Board Configuration binary"
perl: warning: Setting locale failed.
perl: warning: Please check that your locale settings:
        LC_ALL = "zh_CN.UTF-8",
        LC_CTYPE = "zh_CN.UTF-8",
        LANG = "zh_CN.UTF-8"
    are supported and installed on your system.
perl: warning: Falling back to the system default locale ("Chinese (Simplified)_China.936").
Load address not computed for file that does not have a segment table at XML_TI_OFD.pm line 1441.
Use of uninitialized value in multiplication (*) at script/mkhex4bin.pl line 295.
"Sign Board Configuration with GP key"
"  (note: GP board configuration not ACTUALLY signed with anything)"
"Generate include-able header file from signed Board Configuration"
"Build Board Configuration Obj Files"
"Create unsigned Board Configuration binary"
perl: warning: Setting locale failed.
perl: warning: Please check that your locale settings:
        LC_ALL = "zh_CN.UTF-8",
        LC_CTYPE = "zh_CN.UTF-8",
        LANG = "zh_CN.UTF-8"
    are supported and installed on your system.
perl: warning: Falling back to the system default locale ("Chinese (Simplified)_China.936").
Load address not computed for file that does not have a segment table at XML_TI_OFD.pm line 1441.
Use of uninitialized value in multiplication (*) at script/mkhex4bin.pl line 295.
"Sign Board Configuration with GP key"
"  (note: GP board configuration not ACTUALLY signed with anything)"
"Generate include-able header file from signed Board Configuration"
SYSFW Boardcfg blob created at C:\ti\mcu_plus_sdk_am62ax_09_01_00_39\source\drivers\sciclient\sciclient_default_boardcfg\am62ax\boardcfg_blob.bin
SYSFW Boardcfg blob created at C:\ti\mcu_plus_sdk_am62ax_09_01_00_39\source\drivers\sciclient\sciclient_default_boardcfg\am62ax\boardcfg_sbldata_blob.bin

您好、Prashant

我尝试两次、一次不报告错误、但它根据输出 TISCI 跟踪、  

[sbl uart uniflash stage1] recv the sbl_uart_uniflash stage2 image...
[sbl uart uniflash stage1] recv the sbl_uart_uniflash stage2 image Done!!!
[Bootloader:Bootloader_parseMultiCoreAppImage] HS Device : VerifyMulticoreImage
[Bootloader:Bootloader_verifyMulticoreImage] The bootMedia is 0xb0070001
[Bootloader:Bootloader_verifyMulticoreImage] certLoadAddr = 0x84000000
[DM:SCICLIENT] Request Sciclient_service for msgType = 0xc120...
[DM:SCICLIENT] Request Sciclient_serviceSecureProxy for msgType = 0xc120...
[DM:SCICLIENT:Sciclient_procBootAuthAndStart] retVal = 0x0,respParam.flags = 0x2
[Bootloader:Bootloader_verifyMulticoreImage] authStatus = 0x0
[sbl uart uniflash stage1] App_loadSelfcoreImage function : Bootloader_parseMultiCoreAppImage return status 0...
[DM:SCICLIENT] Request Sciclient_service for msgType = 0x10c...
0x6180010C
0x61C00079
0x612F7CA7
0x612F7CA7
[DM:SCICLIENT] Request Sciclient_service for msgType = 0x10e...
0x6180010E
0x61C00079
[sbl uart uniflash stage1] Loading the sbl_uart_uniflash_stage2 image to WKP core...
[sbl uart uniflash stage1] Loading the sbl_uart_uniflash_stage2 image Done and Response message for XMODEM...


[sbl uart uniflash stage2] Init has been Done!!!
ERROR: Board_flashOpen:201: FLASH open failed for instance 0 !!!

其中一个报告身份验证错误、它输出 TISCI 跟踪、如下所示

[sbl uart uniflash stage1] recv the sbl_uart_uniflash stage2 image...
[sbl uart uniflash stage1] recv the sbl_uart_uniflash stage2 image Done!!!
[Bootloader:Bootloader_parseMultiCoreAppImage] HS Device : VerifyMulticoreImage
[Bootloader:Bootloader_verifyMulticoreImage] The bootMedia is 0xb0070001
[Bootloader:Bootloader_verifyMulticoreImage] certLoadAddr = 0x84000000
[DM:SCICLIENT] Request Sciclient_service for msgType = 0xc120...
[DM:SCICLIENT] Request Sciclient_serviceSecureProxy for msgType = 0xc120...
[DM:SCICLIENT:Sciclient_procBootAuthAndStart] retVal = 0x0,respParam.flags = 0x0
[DM:SCICLIENT:Sciclient_procBootAuthAndStart] Sciclient_procBootAuthAndStart request is NOACK
[Bootloader:Bootloader_verifyMulticoreImage] authStatus = 0xffffffff
ERROR: Bootloader_verifyMulticoreImage:658: Failed to authenticate Image
[sbl uart uniflash stage1] App_loadSelfcoreImage function : Bootloader_parseMultiCoreAppImage return status -1...
[sbl uart uniflash stage1] Loading the sbl_uart_uniflash_stage2 image to WKP core...
[sbl uart uniflash stage1] Loading the sbl_uart_uniflash_stage2 image Done and Response message for XMODEM...

请帮助 anaysis,谢谢

BR

Tom

您好、Tom、

如果启用了 TIFS 日志、将按此处所述从存储器中收集

https://software-dl.ti.com/tisci/esd/latest/4_trace/trace.html#trace-memory-buffer-location

此外、如果使用 HSFS 器件类型、请确保将 DEVICE_TYPE 设置为仅 GP、因为 DEVICE_TYPE=HS 适用于 HSSE 器件。

此致、

Prashant

您好、Tom、

Unknown 说：

启用 TIFS 日志时、从唤醒 UART0输出的某些 TIFS 日志中、请参阅

这些是 DM 日志。 TIFS 日志将独立、并从其定义的存储器缓冲区中收集。

我希望您在身份验证失败后以无限循环的方式捕获执行、并收集 TIFS 日志。

什么从存储器缓冲区收集 TIFS 日志？

Unknown 说：

从内存缓冲区收集 TIFS 日志的内容是什么？

您可以使用任何调试器收集日志。

或者、您可以使用以下函数转储它们

void dump_tifs_logs()
{
    #define TIFS_LOGS_BUFFER_ADDR 0x4405F000
    #define TIFS_LOGS_BUFFER_SIZE 0x1000

    uint8_t* ptr = (uint8_t*)TIFS_LOGS_BUFFER_ADDR;

    DebugP_log("\r\n<<TIFS_LOGS\r\n");

    for(int32_t i = 0; i < TIFS_LOGS_BUFFER_SIZE; i++)
    {
        DebugP_log("%c", *ptr);
        ptr++;
    }

    DebugP_log("\r\nTIFS_LOGS\r\n");
}

您好、Tom、

感谢您发送编修。

在解析日志时、我们看到了以下内容

0x0042C120：基端口：TISCI_MSG_RECEIVED (已处理 TISCI 消息中断)：队列 ID：2消息 ID：C120
0x00820024：基端口：TISCI_MSG_SENSOR_HOST_ID (收到来自安全主机的消息)：队列 ID：2主机 ID：36
0x20C00004：安全性：sec_boot (安全引导 API 调用期间的故障点)：0x01 =>证书长度>ASN1P_IMAX、0x02 =>发出获取证书、0x3 =>哈希操作有问题、0x4 =>哈希比较失败：0

失败的原因似乎是散列比较、可能是映像损坏所致。

我认为这主要是一个 DDR 问题。 该映像在请求身份验证之前首先读入 DDR。 因此、如果您的 DDR 配置不稳定、这可能会导致映像有时被错误读取、从而导致身份验证失败。 即使单个字节损坏也足以应对该故障。

那么、问题是、您是否对 DDR 进行了任何应力测试？

此致、

Prashant

您好、Tom、

我们先着重介绍一下 DDR 器件。 在深入研究 DDR 的压力测试之前、您能试一下以下内容吗？

假设身份验证`sbl_uart_uniflash_stage2`、应用修改 stage1 linker.cmd 的以下补丁、以将读取的映像存储在身份验证后的加载地址不同的位置。

diff --git a/examples/drivers/boot/sbl_uart_uniflash_multistage/sbl_uart_uniflash_stage1/am62ax-sk/r5fss0-0_nortos/ti-arm-clang/linker.cmd b/examples/drivers/boot/sbl_uart_uniflash_multistage/sbl_uart_uniflash_stage1/am62ax-sk/r5fss0-0_nortos/ti-arm-clang/linker.cmd
index 31088ce..d9aabe0 100644
--- a/examples/drivers/boot/sbl_uart_uniflash_multistage/sbl_uart_uniflash_stage1/am62ax-sk/r5fss0-0_nortos/ti-arm-clang/linker.cmd
+++ b/examples/drivers/boot/sbl_uart_uniflash_multistage/sbl_uart_uniflash_stage1/am62ax-sk/r5fss0-0_nortos/ti-arm-clang/linker.cmd
@@ -73,5 +73,5 @@ MEMORY
     R5F_TCMB0: ORIGIN = 0x41010000 , LENGTH = 0x00008000 */
     HSM_RAM_VECS: ORIGIN = 0x43C00000 , LENGTH = 0x100
     HSM_RAM  : ORIGIN = 0x43C00100 , LENGTH = 0x3c800 - 0x100
-    DDR      : ORIGIN = 0x84000000 , LENGTH = 0x800000
+    DDR      : ORIGIN = 0x88000000 , LENGTH = 0x800000
 }

这应该读取位于0x88000000的映像、然后请求身份验证。 如果身份验证成功、则会在不同的地址0x84000000处加载最终映像。

因此、想法是在身份验证失败发生时、可能使用调试器将读取的映像保存在0x88000000、并将该映像与 PC 中可用的黄金映像进行比较。 如果它们不同、它将确认解除武装、复员和重返社会是这里的问题、否则我们将不得不去别处看看。

此致、

Prashant

为了防止出现这种情况、还可以通过其他方法来确认图像是否已正确读取。

您可以在应用中集成一个基于软件的简单 CRC32计算。 然后、就在身份验证请求之前、计算 DDR 中读取的映像的 CRC32校验和。 如果身份验证失败、请将计算出的 CRC32与黄金映像的 CRC32进行比较。

我们没有 CRC32代码、但您应该可以在互联网上找到它。 如果找到、我认为这种方法比上一种方法更好。

当使用 UART uniflash 程序时、xds110不连接板、因此没有工具可以转储 DDR 数据。

Unknown 说：

当使用 UART uniflash 程序时、xds110未连接电路板、因此没有工具可以转储 DDR 数据。

我无法理解。 您的电路板是否没有任何外部或板载调试器？ 如果您这样做、您可以在故障后连接到内核并从 DDR 读取数据。

该程序基于串行端口刷写程序、且刷写程序在 DM-R5F 内核上运行、由于该芯片是 FS_HS 芯片、每个引导都需要等待 SYSFW 固件、所以在开发 SBL 时您不能使用 XDS110调试器、不能看到板上没有调试接口。

因此、请勿通过 XDS110转储原始数据

这是不正确的。 在引导流程中、您可以随时连接到 DM-R5F 内核。

我希望您在身份验证失败时立即在无限循环中捕获执行。 然后、使用调试器连接到 DM-R5F 内核并转储来自 DDR 的映像。

您好、Tom、

Unknown 说：

事实证明、DDR 数据已被篡改、下一步是进行 DDR 压力测试

如果确实是这种情况、您能否为 DDR 应力测试创建一个单独的线程、以便将其分配给合适的专家。

DDR 应力测试完成后、我们可以回到这一点、看看问题是否仍然存在。

此致、

Prashant

您好、Prashant

我们使用两块电路板做 DDR 应力测试2 dyas、没有任何错误。

后面的测试日志

Prashant、您好！

我让客户在整个周末在两块有问题的主板上运行 DDR 压力测试仪(他们只有两块主板、都有此问题)。 结果表明、DDR 应力测试没有失败。

这超出了我们的预期、我们不知道下一步将要努力做什么。

谢谢！

Kevin

您好、Tom、

您是否使用了用于 DDR 应力测试的 MCU+ SDK 中的确切 DDR 配置？

此外、 gUniflashFileBuf_bk 数组如何获取其值？ 也许您可以共享将有所帮助的完整 bootloader.c 源文件。

此致、

Prashant

您好、Prashant  

后面的 main.c 代码  

 /*
 *  Copyright (C) 2018-2021 Texas Instruments Incorporated
 *
 *  Redistribution and use in source and binary forms, with or without
 *  modification, are permitted provided that the following conditions
 *  are met:
 *
 *    Redistributions of source code must retain the above copyright
 *    notice, this list of conditions and the following disclaimer.
 *
 *    Redistributions in binary form must reproduce the above copyright
 *    notice, this list of conditions and the following disclaimer in the
 *    documentation and/or other materials provided with the
 *    distribution.
 *
 *    Neither the name of Texas Instruments Incorporated nor the names of
 *    its contributors may be used to endorse or promote products derived
 *    from this software without specific prior written permission.
 *
 *  THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS AND CONTRIBUTORS
 *  "AS IS" AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT
 *  LIMITED TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR
 *  A PARTICULAR PURPOSE ARE DISCLAIMED. IN NO EVENT SHALL THE COPYRIGHT
 *  OWNER OR CONTRIBUTORS BE LIABLE FOR ANY DIRECT, INDIRECT, INCIDENTAL,
 *  SPECIAL, EXEMPLARY, OR CONSEQUENTIAL DAMAGES (INCLUDING, BUT NOT
 *  LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS OR SERVICES; LOSS OF USE,
 *  DATA, OR PROFITS; OR BUSINESS INTERRUPTION) HOWEVER CAUSED AND ON ANY
 *  THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT LIABILITY, OR TORT
 *  (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY OUT OF THE USE
 *  OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF SUCH DAMAGE.
 */

/*
 * Auto generated file - DO NOT MODIFY
 */

#include <stdlib.h>
#include <string.h>
#include "ti_drivers_config.h"
#include "ti_drivers_open_close.h"
#include "ti_board_open_close.h"
#include <drivers/soc.h>
#include <drivers/pinmux.h>
#include <drivers/device_manager/sciclient.h>
#include <drivers/bootloader.h>
#include <drivers/bootloader/bootloader_xmodem.h>
#include <drivers/bootloader/bootloader_uniflash.h>
#include <kernel/dpl/DebugP.h>
#include <drivers/device_manager/sciclient.h>
#include <drivers/sciclient/include/tisci/am62ax/tisci_hosts.h>
#include <drivers/sciclient/include/tisci/am62ax/tisci_boardcfg_constraints.h>
#include <drivers/sciclient/include/tisci/am62ax/tisci_devices.h>

#define BOOTLOADER_UNIFLASH_MAX_FILE_SIZE (0x800000) /* This has to match the size of DDR section in linker.cmd */
uint8_t gUniflashFileBuf[BOOTLOADER_UNIFLASH_MAX_FILE_SIZE] __attribute__((aligned(128), section(".bss.filebuf")));
uint8_t gUniflashFileBuf_bk[BOOTLOADER_UNIFLASH_MAX_FILE_SIZE] __attribute__((aligned(128), section(".bss.filebuf")));

#define BOOTLOADER_UNIFLASH_VERIFY_BUF_MAX_SIZE (32*1024)
uint8_t gUniflashVerifyBuf[BOOTLOADER_UNIFLASH_VERIFY_BUF_MAX_SIZE] __attribute__((aligned(128), section(".bss")));

extern const struct tisci_boardcfg gBoardConfigLow_debug;

/* call this API to stop the booting process and spin, do that you can connect
 * debugger, load symbols and then make the 'loop' variable as 0 to continue execution
 * with debugger connected.
 */
void loop_forever()
{
    volatile uint32_t loop = 1;
    while(loop)
        ;
}

int32_t App_loadSelfcoreImage(Bootloader_Handle bootHandle, Bootloader_BootImageInfo *bootImageInfo)
{
	int32_t status = SystemP_FAILURE;

    if(bootHandle != NULL)
    {
        status = Bootloader_parseMultiCoreAppImage(bootHandle, bootImageInfo);
        DebugP_log("[sbl uart uniflash stage1] App_loadSelfcoreImage function : Bootloader_parseMultiCoreAppImage return status %d...\r\n",status);

        if(status == SystemP_SUCCESS)
        {
            /* Set clocks for self cluster */
            bootImageInfo->cpuInfo[CSL_CORE_ID_R5FSS0_0].clkHz = Bootloader_socCpuGetClkDefault(CSL_CORE_ID_R5FSS0_0);

            /* Reset self cluster, both Core0 and Core 1. Init RAMs and load the app  */
            status = Bootloader_loadSelfCpu(bootHandle, &(bootImageInfo->cpuInfo[CSL_CORE_ID_R5FSS0_0]));
            //status = Bootloader_setSelfCpuEntryPoint(bootHandle, &(bootImageInfo->cpuInfo[CSL_CORE_ID_R5FSS0_0]));
        }
    }

    return status;
}

int main()
{
    int32_t status = SystemP_SUCCESS;
    uint32_t done = 0U;
    uint32_t fileSize;
    Bootloader_UniflashResponseHeader respHeader;
    // Sciclient_BoardCfgPrms_t        sciclient_boardcfg;

    Bootloader_socWaitForFWBoot();
    status = Bootloader_socOpenFirewalls();

    DebugP_assertNoLog(status == SystemP_SUCCESS);

    System_init();
    Board_init();
    Drivers_open();

    status = Board_driversOpen();
    DebugP_assert(status == SystemP_SUCCESS);

    Sciclient_getVersionCheck(1);  

    while(!done)
    {
        /* Xmodem Receive */
        status = Bootloader_xmodemReceive(CONFIG_UART0, gUniflashFileBuf - sizeof(Bootloader_UniflashResponseHeader), BOOTLOADER_UNIFLASH_MAX_FILE_SIZE, &fileSize);
        DebugP_log("[sbl uart uniflash stage1] recv the sbl_uart_uniflash stage2 image...\r\n");
        /*
         * The `fileSize` wouldn't be the actual filesize, but (actual filesize + size of the header + padding bytes) added by xmodem.
         * This adds ~1KB. We can't know exactly how many bytes will be padded without checking the file header. But doing that
         * will unnecessary complicate the logic, so since the overhead is as small as ~1KB we could check for file size exceed
         * by checking * this `fileSize` returned by xmodem as well.
        */
        if(fileSize >= BOOTLOADER_UNIFLASH_MAX_FILE_SIZE)
        {
            /* Possible overflow, send error to host side */
            status = SystemP_FAILURE;

            respHeader.magicNumber = BOOTLOADER_UNIFLASH_RESP_HEADER_MAGIC_NUMBER;
            respHeader.statusCode = BOOTLOADER_UNIFLASH_STATUSCODE_FLASH_ERROR;

            Bootloader_xmodemTransmit(CONFIG_UART0, (uint8_t *)&respHeader, sizeof(Bootloader_UniflashResponseHeader));   
        }

        uint8_t sum = 0x00;
        for(int32_t i = 0; i < BOOTLOADER_UNIFLASH_MAX_FILE_SIZE; i++)
        // for(int32_t i = 0; i < 0x19d8; i++)
        {
            sum = sum + gUniflashFileBuf[i];
            gUniflashFileBuf_bk[i] = gUniflashFileBuf[i];
            // DebugP_log("index = %d,gUniflashFileBuf = 0x%x,gUniflashFileBuf_bk = 0x%x ...\r\n", i,gUniflashFileBuf[i], gUniflashFileBuf_bk[i]);
        }
        
        DebugP_log("[sbl uart uniflash stage1] fileSize is %d...\r\n",fileSize);


        if(status == SystemP_SUCCESS)
        {
            Bootloader_BootImageInfo bootImageInfo;
            Bootloader_Params bootParams;
            Bootloader_Handle bootHandle;

            Bootloader_Params_init(&bootParams);
            Bootloader_BootImageInfo_init(&bootImageInfo);

            bootParams.memArgsAppImageBaseAddr = (uintptr_t)(gUniflashFileBuf);

            bootHandle = Bootloader_open(CONFIG_BOOTLOADER_MEM, &bootParams);

            DebugP_log("[sbl uart uniflash stage1] recv the sbl_uart_uniflash stage2 image Done!!!\r\n");

            if((bootHandle != NULL) && (SystemP_SUCCESS == status))
            {
                status = App_loadSelfcoreImage(bootHandle, &bootImageInfo);
                DebugP_log("[sbl uart uniflash stage1] Loading the sbl_uart_uniflash_stage2 image to WKP core...\r\n");

            }

            // if(SystemP_SUCCESS == status)  //Has Issue in the A0 Sample
            {
                DebugP_log("[sbl uart uniflash stage1] Loading the sbl_uart_uniflash_stage2 image Done and Response message for XMODEM...\r\n");
                respHeader.magicNumber = BOOTLOADER_UNIFLASH_RESP_HEADER_MAGIC_NUMBER;
                respHeader.statusCode = BOOTLOADER_UNIFLASH_STATUSCODE_SUCCESS;
                status = Bootloader_xmodemTransmit(CONFIG_UART0, (uint8_t *)&respHeader, sizeof(Bootloader_UniflashResponseHeader));
            }

            if(status == SystemP_SUCCESS)
            {
                Bootloader_close(bootHandle);
                done = TRUE;
            }

        }
    }

    /* Call DPL deinit to close the tick timer and disable interrupts before jumping to Stage2*/
    Dpl_deinit();

    Bootloader_JumpSelfCpu();

    Drivers_close();
    Board_deinit();
    System_deinit();

    return 0;
}

未经修改的 SDK 中的 bootloade.c。

在 MCU+SDK 中、不运行 DDR 应力测试、而是针对 Linux 测试内核。  

在 MCU+SDK 中、未修改任何使用 SDK 默认配置的 DDR 配置

您好、Tom、

Unknown 说：

并且在 MCU+SDK 中未修改任何使用 SDK 默认配置的 DDR 配置

我假设您的定制电路板上有定制的 DDR。 如果是、您需要为 DDR 配置正确的 DDR 配置、并将它们集成在 MCU+ SDK 中。

在运行所有通过的压力测试之前、您是否在 Linux 中集成了正确的 DDR 配置？

您好、Prashant

我们使用的 DDR 与 EVM 相同。 我认为 SDK 具有 DDR 配置、我们没有修改。

BR

Tom

您好、Tom、

请允许我有一天或两天再次回顾整个回答、看看是否有任何遗漏。

感谢您的耐心。

您好、Prashant

请提供任何线索

BR

Tom

您好、Tom、

您当前使用的是 MCU+ SDK v9.1。 您能否尝试一下最新发布的 v9.2？

https://www.ti.com/tool/download/MCU-PLUS-SDK - AM62A/09.02.00.38

此致、

Prashant