Linux内核模块作弊-Qemu Baremetal Xilinx Zynq A9 [英] Linux Kernel Module Cheat - Qemu Baremetal Xilinx Zynq A9
问题描述
我的目标是模拟运行裸机软件的Zynq-7000上的ARM A9处理器.我已经尝试了两种不同的方法来解决这两个问题.有关如何进行的任何建议将不胜感激.
My goal is to emulate the ARM A9 processor as found on the Zynq-7000, running baremetal software. I have tried 2 different approaches to this and run into road blocks on both. Any suggestions on how to proceed would be appreciated.
Linux内核模块作弊(LKMC，使用v3.0)
使用 ./build --arch arm qemu-baremetal
which links to
Linux Kernel Module Cheat (LKMC, using v3.0)
built using ./build --arch arm qemu-baremetal
使用ARM虚拟机( -virt 标志)的站点上的示例正常工作.尝试修改此设置以使其适合我的设置是导致问题的原因(以下详细信息).
The examples on the site of using the ARM virtual machine (-virt flag) work fine. Trying to modify this to work with my setup is what's causing problems (details below).
我尝试复制示例命令行调用，但是使用了 -cpu cortex-a9 选项: qemu-system-arm:mach-virt:不支持CPU cortex-a9
然后我将整个调用更改为 qemu-system-arm -M xilinx-zynq-a9 -cpu cortex-a9 -nographic -kernel hello.elf -m 512M -serial mon:stdio -s -S
它崩溃并显示错误 qemu:致命的:试图在RAM或ROM外部以0x40000000执行代码
这是有道理的，因为该应用程序是使用LKMC构建的，而我试图在该框架之外运行它.
I tried to copy the example command line invocation, but with the -cpu cortex-a9 option instead:
qemu-system-arm: mach-virt: CPU cortex-a9 not supported
Then I changed the whole invocation to be
qemu-system-arm -M xilinx-zynq-a9 -cpu cortex-a9 -nographic -kernel hello.elf -m 512M -serial mon:stdio -s -S
And it crashed with the error
qemu: fatal: Trying to execute code outside RAM or ROM at 0x40000000
Which makes sense, because the application was built with the LKMC, and I was trying to run it outside of that framework.
因此，我然后尝试运行自己的应用程序，该应用程序是使用Xilinx工具链的修改版进行编译的.我确信它不会立即生效，因为启动顺序中的某些部分我必须更改.但是我试图弄清楚它们是什么并进行更改.与
一起运行 qemu-system-arm -M xilinx-zynq-a9 -cpu cortex-a9 -nographic -kernel helloworld.elf -m 512M -serial mon:stdio -s -S
允许GDB成功连接，但无法正确读取符号表.使用 arm-none-eabi-objdump -D helloworld.elf
告诉我 main 位于 0x001004fc ，但是GDB认为它位于 0x40000324 (使用命令 info address main )
So then I tried running my own application, which was compiled using a modified version of the Xilinx toolchain. I'm sure it won't work right away, as there will be some parts of the bootup sequence that I have to change. But I'm trying to figure out what those are and change them.
Running with
qemu-system-arm -M xilinx-zynq-a9 -cpu cortex-a9 -nographic -kernel helloworld.elf -m 512M -serial mon:stdio -s -S
allows GDB to connect successfully, but it can't read the symbol table properly. Using
arm-none-eabi-objdump -D helloworld.elf
tells me that main is at 0x001004fc, but GDB thinks it's at 0x40000324 (using the command info address main).
PYNQ-Z1(网页， qemu-system-arm 而不是 qemu-system-aarch64 的原因.如果那是错误的，有人可以纠正我.
The PYNQ-Z1 (webpage, datasheet) has a 32-bit ARM Cortex-A9 processor, so that's why I'm using qemu-system-arm instead of qemu-system-aarch64. Someone can correct me if that's wrong.
请注意，我不能简单地切换到其他体系结构.我使用的代码不能容忍更改，除了进行细微调整外，不能使板级支持程序包(BSP)与模拟器兼容，而不会损害我的研究的有效性.
As a note, I cannot simply switch to a different architecture; the code that I am using cannot tolerate changes other than small tweaks to make the Board Support Package (BSP) compatible with the simulator, without harming the validity of my research.
我已经离开了一段时间了 ./run --arch arm -m 512M --baremetal pynq/helloworld --wait-gdb ./run-gdb --arch arm --baremetal pynq/helloworld --no-continue-main
然后我逐步使用GDB查找数据中止的位置，并找出Qemu不支持哪种硬件.
What I have been going off of for a while now is
./run --arch arm -m 512M --baremetal pynq/helloworld --wait-gdb
./run-gdb --arch arm --baremetal pynq/helloworld --no-continue -- main
and I step through using GDB to find where there are data aborts and figure out what kind of hardware is not supported by Qemu.
我正在运行的软件是使用经过修改的Xilinx工具链构建的，因此包含许多Xilinx标准库功能.到目前为止，在修改代码以使其与虚拟机一起使用时，我发现了一些更改，例如更改UART设备的地址和禁用一些启动任务，例如使SCU无效或更改缓存控制器配置，这可能是因为这些东西不是Qemu模仿的.
The software that I am running is built using a modified Xilinx toolchain, and so includes many of the Xilinx standard library functions. In modifying the code to work with the virtual machine, I have discovered a few changes so far, such as changing the address of the UART device and disabling some boot-up tasks such as invalidating the SCU or changing the cache controller configuration, presumably because these things are not emulated by Qemu.
调试启动时，启动时遇到的下一个问题是XTime函数(xtime_l.c).这些函数是读取全局系统计时器的包装器.Qemu界面中的命令 info mtree 的结果似乎表明没有与之交互的全局计时器设备.是否可以将计时器设备添加到ARM虚拟机?只要能以与Zynq相同的方式使用寄存器读和写，就可以使用基地址，这无关紧要.
When debugging bootup, the next problem I have run into booting up is the XTime functions (xtime_l.c). These functions are wrappers around reading the global system timer. The results of the command info mtree in the Qemu interface seem to indicate that there is no global timer device with which to interact. Is there a way to add a timer device to the ARM virtual machine? It doesn't matter what the base address is, as long as it can be used in the same way as on the Zynq, using register reads and writes.
然后我尝试使用特定的机器标志 xilinx-zynq-a9 .LKMC生成以下命令:
Then I tried to use the specific machine flag xilinx-zynq-a9. LKMC generates the following command:
+ /home/$USER/$INSTALL_DIR/out/qemu/default/arm-softmmu/qemu-system-arm \
-machine xilinx-zynq-a9 \
-gdb tcp::45457 \
-kernel /home/$USER/$INSTALL_DIR/out/baremetal/arm/qemu/xilinx-zynq-a9/hello.elf \
-m 512M \
-monitor telnet::45454,server,nowait \
-netdev user,hostfwd=tcp::45455-:45455,hostfwd=tcp::45456-:22,id=net0 \
-no-reboot \
-smp 1 \
-virtfs local,path=/home/$USER/$INSTALL_DIR/data/9p,mount_tag=host_data,security_model=mapped,id=host_data \
-virtfs local,path=/home/$USER/$INSTALL_DIR/out,mount_tag=host_out,security_model=mapped,id=host_out \
-virtfs local,path=/home/$USER/$INSTALL_DIR/rootfs_overlay,mount_tag=host_rootfs_overlay,security_model=mapped,id=host_rootfs_overlay \
-serial mon:stdio \
-trace enable=load_file,file=/home/$USER/$INSTALL_DIR/out/run/qemu/arm/0/trace.bin \
-cpu cortex-a9 \
-device virtio-gpu-pci \
-nographic \
-serial tcp::45458,server,nowait \
-semihosting \
;
此虚拟机与通用虚拟机之间的唯一区别是指定计算机和cpu的行，它们分别是 -machine virt -machine highmem = off 和 -cpu cortex-a15 (实际上我必须修改LKMC代码，以使其输出机器的正确cpu名称).
The only differences between this and the generic virtual machine are the lines that specify the machine and the cpu, which used to be -machine virt -machine highmem=off and -cpu cortex-a15 respectively (I actually had to modify the LKMC code to get it to output the correct cpu name for the machine).
但是，此操作失败，并显示错误 qemu-system-arm:-device rtl8139，netdev = net0:未找到设备'rtl8139'的'PCI'总线
这是有道理的，因为并非所有Zynq部件都具有PCI总线.因此，大多数情况下，我想知道为什么当目标仍然是裸机时，LKMC为什么会生成这样的命令序列.
However, this fails with the error
qemu-system-arm: -device rtl8139,netdev=net0: No 'PCI' bus found for device 'rtl8139'
This makes sense, because not all Zynq parts have PCI buses. So mostly I am wondering why LKMC would generate such a sequence of commands when the target is baremetal anyway.
我认为第一个选项最有可能起作用，因为似乎 -virt 机器比某些特定目标具有更好的支持.有趣的是Xilinx SDK随附的Qemu版本不支持Zynq的裸机(在
The first option I think is the most likely to work, since it seems like the -virt machine has better support than some of the specific targets. It is interesting that the version of Qemu that ships with the Xilinx SDK does not support baremetal with the Zynq (referred to as "standalone" in the Xilinx Docs).
是否可以将计时器设备添加到ARM虚拟机?
有没有人在Qemu Xilinx ARM A9上运行裸机代码?
Is there a way to add a timer device to the ARM virtual machine?
Has anyone run baremetal code on Qemu Xilinx ARM A9?
我尽力做到尽可能具体，但随时可以提出澄清的问题.
I've tried to be as specific as possible, but feel free to ask clarifying questions.
推荐答案
- 与添加对计时器的支持相比，这可能是一项艰巨的任务设备安装到现有QEMU机器上.更具体地说，这可能不是之所以需要它，是因为它们中的相当一部分都支持ARM.体系结构计时器或特定的计时器硬件.
在xilinx-zynq-a9的特定情况下，似乎Global从Zynq-7000技术支持参考手册. - 几次阅读您的文章后，我得出的结论是，您正在使用的一系列工具(KMC，工具链，QEMU)可能会出错.因此，我希望创建一个与QEMU xilinx-一起使用的裸机应用程序的最小，可重现的示例.我确实信任使用arm工具链的zynq-a9机器，以及使用我编写的脚本从头开始构建的QEMU的最新版本4.2.0.
- This is probably a non-trivial task than to add support for a timer
device to an existing QEMU machine. More specifically, this may not
be needed since a fair amount of them either support an ARM
architectural timer or a specific timer hardware.
In the specific case of the xilinx-zynq-a9, it seem theGlobal Timer Counterdescribed from page 1448 of the Zynq-7000 Technical Reference Manual is supported. - After having reading your post a couple of times, I reached the conclusion that a lot of things may go wrong with the set of tools you are using (KMC, toolchain, QEMU). I therefore created what I hope is a Minimal, Reproducible Example of a bare-metal application working with a QEMU xilinx-zynq-a9 machine using an arm toolchain I do trust, and the latest version of QEMU, 4.2.0, built from scratch using a script I wrote.
请注意，为了回答您的问题，我改编了我已经可以使用的现有个人项目，并且我知道它正在工作.
Please note that I adapted an existing personal project I already had available, and I know is working, for the purpose of answering your question.
构建QEMU:执行 build-qemu.sh -此脚本可在64位Ubuntu 18.04和19.10上运行，您必须将 PERL_MODULES_VERSION 设置为 5.28 .
Building QEMU: execute build-qemu.sh - this script works on 64 bits Ubuntu 18.04 and 19.10, you will have to set PERL_MODULES_VERSION to 5.28.
build-qemu.sh :
#!/bin/bash
set -e
QEMU_VERSION=4.2.0
# xenial
PERL_MODULES_VERSION=5.22
# eoan
PERL_MODULES_VERSION=5.28
# bionic
PERL_MODULES_VERSION=5.26
PREFIX=/opt/qemu-${QEMU_VERSION}
do_install_prerequisites()
{
sudo apt-get install libglib2.0-dev libfdt-dev libpixman-1-dev zlib1g-dev libaio-dev libbluetooth-dev libbrlapi-dev libbz2-dev libcap-dev libcap-ng-dev libcurl4-gnutls-dev libgtk-3-dev libibverbs-dev \
libjpeg8-dev libncurses5-dev libnuma-dev librbd-dev librdmacm-dev libsasl2-dev libsdl2-dev libseccomp-dev libsnappy-dev libssh2-1-dev libvde-dev libvdeplug-dev libvte-2.91-dev libxen-dev liblzo2-dev \
valgrind xfslibs-dev liblzma-dev flex bison texinfo perl perl-modules-${PERL_MODULES_VERSION} python-sphinx gettext
}
do_download_qemu()
{
if [ ! -f qemu-${QEMU_VERSION}.tar.xz ]
then
wget https://download.qemu.org/qemu-${QEMU_VERSION}.tar.xz
fi
}
do_extract_qemu()
{
echo "extracting..."
rm -rf qemu-${QEMU_VERSION}
tar Jxf qemu-${QEMU_VERSION}.tar.xz
}
do_configure_qemu()
{
local TARGET_LIST="arm-softmmu"
pushd qemu-${QEMU_VERSION}
./configure --target-list="${TARGET_LIST}" --prefix=${PREFIX} --extra-cflags="-I$(pwd)/packages/include" --extra-ldflags="-L$(pwd)/packages/lib"
popd
}
do_build_qemu()
{
echo "building..."
pushd qemu-${QEMU_VERSION}
make all
popd
}
do_install_qemu()
{
echo "installing..."
pushd qemu-${QEMU_VERSION}
sudo make install
popd
}
do_build()
{
do_download_qemu
do_extract_qemu
do_configure_qemu
do_build_qemu
do_install_qemu
}
# main
do_install_prerequisites
do_build
脚本完成后，您应该已安装qemu-system-user 4.2.0:
Upon the script completion, you should have qemu-system-user 4.2.0 installed:
ls -gG /opt/qemu-4.2.0/bin/
total 22992
-rwxr-xr-x 1 22520 Feb 21 23:57 elf2dmp
-rwxr-xr-x 1 18424 Feb 21 23:57 ivshmem-client
-rwxr-xr-x 1 218264 Feb 21 23:57 ivshmem-server
-rwxr-xr-x 1 30864 Feb 21 23:57 qemu-edid
-rwxr-xr-x 1 374328 Feb 21 23:57 qemu-ga
-rwxr-xr-x 1 1767744 Feb 21 23:57 qemu-img
-rwxr-xr-x 1 1719104 Feb 21 23:57 qemu-io
-rwxr-xr-x 1 505016 Feb 21 23:57 qemu-keymap
-rwxr-xr-x 1 1727744 Feb 21 23:57 qemu-nbd
-rwxr-xr-x 1 599848 Feb 21 23:57 qemu-pr-helper
-rwxr-xr-x 1 16510840 Feb 21 23:57 qemu-system-arm
-rwxr-xr-x 1 26856 Feb 21 23:57 virtfs-proxy-helper
我们现在需要创建以下文件:
We now need to create the following files:
gcc_arm32_ram.ld (adapted from the standard GCC CMSIS 5.60 linker script):
/******************************************************************************
* @file gcc_arm32.ld
* @brief GNU Linker Script for Cortex-M based device
* @version V2.0.0
* @date 21. May 2019
******************************************************************************/
/*
* Copyright (c) 2009-2019 Arm Limited. All rights reserved.
*
* SPDX-License-Identifier: Apache-2.0
*
* Licensed under the Apache License, Version 2.0 (the License); you may
* not use this file except in compliance with the License.
* You may obtain a copy of the License at
*
* www.apache.org/licenses/LICENSE-2.0
*
* Unless required by applicable law or agreed to in writing, software
* distributed under the License is distributed on an AS IS BASIS, WITHOUT
* WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
* See the License for the specific language governing permissions and
* limitations under the License.
*/
MEMORY
{
RAM (rwx) : ORIGIN = __RAM_BASE, LENGTH = __RAM_SIZE
}
/* Linker script to place sections and symbol values. Should be used together
* with other linker script that defines memory regions FLASH and RAM.
* It references following symbols, which must be defined in code:
* Reset_Handler : Entry of reset handler
*
* It defines following symbols, which code can use without definition:
* __exidx_start
* __exidx_end
* __copy_table_start__
* __copy_table_end__
* __zero_table_start__
* __zero_table_end__
* __etext
* __data_start__
* __preinit_array_start
* __preinit_array_end
* __init_array_start
* __init_array_end
* __fini_array_start
* __fini_array_end
* __data_end__
* __bss_start__
* __bss_end__
* __end__
* end
* __HeapLimit
* __StackLimit
* __StackTop
* __stack
*/
ENTRY(Reset_Handler)
SECTIONS
{
.text :
{
KEEP(*(.vectors))
*(.text*)
KEEP(*(.init))
KEEP(*(.fini))
/* .ctors */
*crtbegin.o(.ctors)
*crtbegin?.o(.ctors)
*(EXCLUDE_FILE(*crtend?.o *crtend.o) .ctors)
*(SORT(.ctors.*))
*(.ctors)
/* .dtors */
*crtbegin.o(.dtors)
*crtbegin?.o(.dtors)
*(EXCLUDE_FILE(*crtend?.o *crtend.o) .dtors)
*(SORT(.dtors.*))
*(.dtors)
*(.rodata*)
KEEP(*(.eh_frame*))
} > RAM
/*
* SG veneers:
* All SG veneers are placed in the special output section .gnu.sgstubs. Its start address
* must be set, either with the command line option �--section-start� or in a linker script,
* to indicate where to place these veneers in memory.
*/
/*
.gnu.sgstubs :
{
. = ALIGN(32);
} > RAM
*/
.ARM.extab :
{
*(.ARM.extab* .gnu.linkonce.armextab.*)
} > RAM
__exidx_start = .;
.ARM.exidx :
{
*(.ARM.exidx* .gnu.linkonce.armexidx.*)
} > RAM
__exidx_end = .;
.copy.table :
{
. = ALIGN(4);
__copy_table_start__ = .;
LONG (__etext)
LONG (__data_start__)
LONG (__data_end__ - __data_start__)
/* Add each additional data section here */
/*
LONG (__etext2)
LONG (__data2_start__)
LONG (__data2_end__ - __data2_start__)
*/
__copy_table_end__ = .;
} > RAM
.zero.table :
{
. = ALIGN(4);
__zero_table_start__ = .;
/* Add each additional bss section here */
/*
LONG (__bss2_start__)
LONG (__bss2_end__ - __bss2_start__)
*/
__zero_table_end__ = .;
} > RAM
/**
* Location counter can end up 2byte aligned with narrow Thumb code but
* __etext is assumed by startup code to be the LMA of a section in RAM
* which must be 4byte aligned
*/
__etext = ALIGN (4);
.data : AT (__etext)
{
__data_start__ = .;
*(vtable)
*(.data)
*(.data.*)
. = ALIGN(4);
/* preinit data */
PROVIDE_HIDDEN (__preinit_array_start = .);
KEEP(*(.preinit_array))
PROVIDE_HIDDEN (__preinit_array_end = .);
. = ALIGN(4);
/* init data */
PROVIDE_HIDDEN (__init_array_start = .);
KEEP(*(SORT(.init_array.*)))
KEEP(*(.init_array))
PROVIDE_HIDDEN (__init_array_end = .);
. = ALIGN(4);
/* finit data */
PROVIDE_HIDDEN (__fini_array_start = .);
KEEP(*(SORT(.fini_array.*)))
KEEP(*(.fini_array))
PROVIDE_HIDDEN (__fini_array_end = .);
KEEP(*(.jcr*))
. = ALIGN(4);
/* All data end */
__data_end__ = .;
} > RAM
/*
* Secondary data section, optional
*
* Remember to add each additional data section
* to the .copy.table above to asure proper
* initialization during startup.
*/
/*
__etext2 = ALIGN (4);
.data2 : AT (__etext2)
{
. = ALIGN(4);
__data2_start__ = .;
*(.data2)
*(.data2.*)
. = ALIGN(4);
__data2_end__ = .;
} > RAM2
*/
.bss :
{
. = ALIGN(4);
__bss_start__ = .;
*(.bss)
*(.bss.*)
*(COMMON)
. = ALIGN(4);
__bss_end__ = .;
} > RAM AT > RAM
/*
* Secondary bss section, optional
*
* Remember to add each additional bss section
* to the .zero.table above to asure proper
* initialization during startup.
*/
/*
.bss2 :
{
. = ALIGN(4);
__bss2_start__ = .;
*(.bss2)
*(.bss2.*)
. = ALIGN(4);
__bss2_end__ = .;
} > RAM2 AT > RAM2
*/
.heap (COPY) :
{
. = ALIGN(8);
__end__ = .;
PROVIDE(end = .);
. = . + __HEAP_SIZE;
. = ALIGN(8);
__HeapLimit = .;
} > RAM
.stack (ORIGIN(RAM) + LENGTH(RAM) - __STACK_SIZE) (COPY) :
{
. = ALIGN(8);
__StackLimit = .;
. = . + __STACK_SIZE;
. = ALIGN(8);
__StackTop = .;
} > RAM
PROVIDE(__stack = __StackTop);
/* Check if data + heap + stack exceeds RAM limit */
ASSERT(__StackLimit >= __HeapLimit, "region RAM overflowed with stack")
}
Makefile.inc
# Shared Makefile
.PHONY: clean
all: $(MACHINE).elf
$(MACHINE).elf: $(SOURCES) $(MACHINE).c
$(CC) $(CFLAGS) $(LDFLAGS) -o $(MACHINE).elf $(MACHINE).c $(SOURCES)
$(OBJDUMP) -d $(MACHINE).elf > $(MACHINE).lst
qemu: $(MACHINE).elf
$(QEMU_SYSTEM) -m 513M -nographic -machine $(MACHINE) $(QEMU_DEBUG_OPTIONS) -cpu $(CPU) -kernel $(MACHINE).elf
gdb: $(MACHINE).elf
$(GDB) --quiet --command=$(GDB_COMMANDS) $(MACHINE).elf
clean:
rm -f $(MACHINE).elf $(MACHINE).lst
startup-aarch32.s :
.title startup-aarch32.s
.arch armv7-a
.text
.section .text.startup,"ax"
.globl Reset_Handler
Reset_Handler:
ldr r0, =__StackTop
mov sp, r0
bl start
wait: wfe
b wait
.end
xilinx-zynq-a9.c :
#include <stdint.h>
/* Reference: https://www.xilinx.com/support/documentation/user_guides/ug585-Zynq-7000-TRM.pdf - page 1449.
1. Read the upper 32-bit timer counter register
2. Read the lower 32-bit timer counter register
3. Read the upper 32-bit timer counter register again.
If the value is different to the32-bit upper value read previously, go back to step 2.
Otherwise the 64-bit timercounter value is correct.
*/
static const uintptr_t Global_Timer_Counter_Register0 = 0xF8F00200;
static const uintptr_t Global_Timer_Counter_Register1 = 0xF8F00204;
void start()
{
uint64_t global_timer_counter = 0;
uint32_t upper = 0;
uint32_t upper2 = 0;
uint32_t lower = 0;
for (;;) {
upper = *(volatile uint32_t*) Global_Timer_Counter_Register1;
lower = *(volatile uint32_t*) Global_Timer_Counter_Register0;
upper2 = *(volatile uint32_t*) Global_Timer_Counter_Register1;
if (upper != upper2) {
lower = *(volatile uint32_t*) Global_Timer_Counter_Register0;
}
global_timer_counter = (uint64_t) upper << 32 | lower;
}
}
xilinx-zynq-a9.gdb :
target remote localhost:1234
monitor reset halt
load
break Reset_Handler
break start
xilinx-zynq-a9.ld :
/*
*-------- <<< Use Configuration Wizard in Context Menu >>> -------------------
*/
/*--------------------- Embedded RAM Configuration ----------------------------
<h> RAM Configuration
<o0> RAM Base Address <0x0-0xFFFFFFFF:8>
<o1> RAM Size (in Bytes) <0x0-0xFFFFFFFF:8>
</h>
-----------------------------------------------------------------------------*/
__RAM_BASE = 0x00100000;
__RAM_SIZE = 0x20000000;
/*--------------------- Stack / Heap Configuration ----------------------------
<h> Stack / Heap Configuration
<o0> Stack Size (in Bytes) <0x0-0xFFFFFFFF:8>
<o1> Heap Size (in Bytes) <0x0-0xFFFFFFFF:8>
</h>
-----------------------------------------------------------------------------*/
__STACK_SIZE = 0x00020000;
__HEAP_SIZE = 0x00080000;
/*
*-------------------- <<< end of configuration section >>> -------------------
*/
INCLUDE gcc_arm32_ram.ld
我指定了一个从0x20000000开始的512MiB DDR RAM区域-有关
I specified a 512MiB of DDR RAM area starting from 0x20000000 - see here for more information on the Zynq-7000 memory map.
xilinx-zynq-a9.mak :
# Toolchain
CROSS_COMPILE=/opt/arm/9/gcc-arm-9.2-2019.12-x86_64-arm-none-eabi/bin/arm-none-eabi-
CC=$(CROSS_COMPILE)gcc
OBJDUMP=$(CROSS_COMPILE)objdump
OBJCOPY=$(CROSS_COMPILE)objcopy
# Target
CPU=cortex-a9
MACHINE=xilinx-zynq-a9
CFLAGS=-O0 -ggdb -mtune=$(CPU) -nostdlib -nostartfiles -ffreestanding
LDFLAGS=-L. -Wl,-T,$(MACHINE).ld
SOURCES=startup-aarch32.s
# qemu
QEMU_DEBUG_OPTIONS=-S -gdb tcp::1234,ipv4
QEMU_SYSTEM=/opt/qemu-4.2.0/bin/qemu-system-arm
# GDB
GDB=$(CROSS_COMPILE)gdb
GDB_COMMANDS=${MACHINE}.gdb
include Makefile.inc
You now need yo install the latest GCC toolchain provided by arm:
wget "https://developer.arm.com/-/media/Files/downloads/gnu-a/9.2-2019.12/binrel/gcc-arm-9.2-2019.12-x86_64-arm-none-eabi.tar.xz?revision=64186c5d-b471-4c97-a8f5-b1b300d6594a&la=en&hash=5E9204DA5AF0B055B5B0F50C53E185FAA10FF625" -o gcc-arm-9.2-2019.12-x86_64-arm-none-eabi.tar.xz
mkdir -p /opt/arm/9
tar Jxf gcc-arm-9.2-2019.12-x86_64-arm-none-eabi.tar.xz -C /opt/arm/9
您现在可以编译/执行/调试示例了:
You are now ready to compile/execute/debug the example:
make -f xilinx-zynq-a9.mak clean all
您应该获得以下输出:
rm -f xilinx-zynq-a9.elf xilinx-zynq-a9.lst
/opt/arm/9/gcc-arm-9.2-2019.12-x86_64-arm-none-eabi/bin/arm-none-eabi-gcc -O0 -ggdb -mtune=cortex-a9 -nostdlib -nostartfiles -ffreestanding -L. -Wl,-T,xilinx-zynq-a9.ld -o xilinx-zynq-a9.elf xilinx-zynq-a9.c startup-aarch32.s
/opt/arm/9/gcc-arm-9.2-2019.12-x86_64-arm-none-eabi/bin/arm-none-eabi-objdump -d xilinx-zynq-a9.elf > xilinx-zynq-a9.lst
您可以启动QEMU:
make -f xilinx-zynq-a9.mak qemu
应显示已执行的QEMU命令:
The QEMU command that was executed should be displayed:
/opt/qemu-4.2.0/bin/qemu-system-arm -m 513M -nographic -machine xilinx-zynq-a9 -S -gdb tcp::1234,ipv4 -cpu cortex-a9 -kernel xilinx-zynq-a9.elf
在另一个Shell中，启动GDB:
In an other shell, start GDB:
make -f xilinx-zynq-a9.mak gdb
您应该看到以下输出:
/opt/arm/9/gcc-arm-9.2-2019.12-x86_64-arm-none-eabi/bin/arm-none-eabi-gdb --quiet --command=xilinx-zynq-a9.gdb xilinx-zynq-a9.elf
Reading symbols from xilinx-zynq-a9.elf...
Reset_Handler () at startup-aarch32.s:7
7 ldr r0, =__StackTop
unknown command: 'reset'
Loading section .text, size 0xd8 lma 0x100000
Loading section .copy.table, size 0xc lma 0x1000d8
Start address 0x1000b8, load size 228
Transfer rate: 222 KB/sec, 114 bytes/write.
Breakpoint 1 at 0x1000b8: file startup-aarch32.s, line 7.
Breakpoint 2 at 0x10000c: file xilinx-zynq-a9.c, line 17.
(gdb)
执行继续:
(gdb) continue
Continuing.
Breakpoint 2, start () at xilinx-zynq-a9.c:17
17 uint64_t global_timer_counter = 0;
现在，执行几个 step 命令，并在每行显示 global_timer_counter 变量:
Now, execute several step commands, and display the global_timer_counter variable everytime line:
31 global_timer_counter = (uint64_t) upper << 32 | lower;
已执行:
(gdb) p/x global_timer_counter
$2 = 0xa6f2a0a
(gdb) p/x global_timer_counter
$9 = 0xa84315b
(gdb) p/x global_timer_counter
$10 = 0xabe77cf
64位变量保持递增¸，这与 QEMU 对Zynq全局计时器计数器的有效仿真一致，现在我们有了一个可以调试的裸机Zynq-7000示例使用GDB.
The 64 bits variable keeps incrementing ¸which is consistant with a working emulation of the Zynq Global Timer Counter by QEMU, and we now have a working bare-metal Zynq-7000 example that can be debugged using GDB.
我希望我回答了你问的两个问题.
I hope I answered to the two questions you asked.
这篇关于Linux内核模块作弊-Qemu Baremetal Xilinx Zynq A9的文章就介绍到这了，希望我们推荐的答案对大家有所帮助，也希望大家多多支持IT屋！
