Remote 'g' packet reply is too long - gdb

I am trying to debug Linux kernel with kvm vm. I am getting an error message "Remote 'g' packet reply is too long". My host is 64-bit and so is my vm.
My steps:
Start the VM with custom -kernel, -initrd and -append options.
Start gdb
Execute "set architecture i386:x86-64:intel"
Execute "add-symbol-file linux-3.0/vmlinux"
Execute "show arch" to verify its still "i386:x86-64:intel"
Execute "target remote localhost:1234"
Execute "continue"
Press Ctrl+C, I get the above message.
Has anyone faced this problem?

gdb doesn't work well against a cpu that switches between instruction sets at runtime. Wait for the kernel to leave early boot before connecting, and don't use qemu's -S flag.

I also faced same issue, I fixed it by modifying gdbstub.c (in qemu sources) to send 64bit registers always and hinting GDB that architecture is 64bit by passing set arch i386:x86-64
You can check the patch here:
Visit [URL no longer available]

I found a similar problem (& this question) connecting gdb very early in the boot process – as mentioned in other answers, gdb does not very much appreciate the size of registers changing out from under it. This problem can be seen by using set debug remote 1:
(gdb) set debug remote 1
(gdb) target remote localhost:1234
Remote debugging using localhost:1234
...
Sending packet: $g#67...Ack
Packet received: 000000000000000... <~600 bytes>
(gdb) until *0x1000 # at this address we'll be in a different cpu mode
...
Sending packet: $g#67...Ack
Packet received: 10000080000000000000000000000000800000c... <~1000 bytes>
...
Remote 'g' packet reply is too long: 1000008000000000000000000...
(gdb)
Patching gdb to resize its internal buffer when it sees a too-large packet
as found on this issue in the gdb bug tracker (and elsewhere), does indeed work around the problem, as does patching QEMU to only send 64-bit-sized packets. However, the latter solution breaks debugging in non-64-bit-modes, and it seems that the former fix could be incomplete:
It sounds quite wrong to be changing the target behind
GDB's back when GDB is already debugging it. Not just the size
of the g/G packets may change inadvertently, but the layout as well.
If the target description changes with your re-configuration, it
sounds to me like GDB should fetch/recompute the whole target
description. Today, I think that can only be done with a
disconnect/reconnect.
– https://sourceware.org/ml/gdb/2014-02/msg00005.html
The disconnect/reconnect workaround mentioned at the end of the post does appear to work:
(gdb) disconnect
Ending remote debugging.
(gdb) set architecture i386:x86-64
The target architecture is assumed to be i386:x86-64
(gdb) target remote localhost:1234
Remote debugging using localhost:1234
(gdb) info registers
rax 0x80000010 2147483664
rbx 0x0 0
...

I had accidentally omitted the binary name as an argument to gdb. So this worked for me.
$ gdb ./vmlinux
(gdb) target remote localhost:1234
And then got the Output :
Remote debugging using localhost:1234
0xffffffff81025f96 in default_idle ()
The debugger needs vmlinux so make sure you provide it. OP has a different problem, But my answer might help to those who forgot to provide argument to gdb and ended up with the same error message as OP.

Related

Flashing and running from clion fails with OpenOCD: Error the tcl port command must be used before init

Okay I have this issue for quite some time now because I never found a solution online.
I thought I will ask here and hope if someone else also encountered this issue (and with a little luck solved it).
So the expected behavior of those buttons starting with the left circled one is that the code is compiled, flashed and runs on the embedded hardware. The second circled one is the same as previous but will listen to breakpoints and start a gdb server.
When pressing the first button it will fail with the following error message:
[0mOpen On-Chip Debugger 0.10.0 (2020-07-29) [https://github.com/sysprogs/openocd]
Licensed under GNU GPL v2
libusb1 09e75e98b4d9ea7909e8837b7a3f00dda4589dc3
For bug reports, read
http://openocd.org/doc/doxygen/bugs.html
WARNING: interface/stlink-v2.cfg is deprecated, please switch to interface/stlin
k.cfg
Info : auto-selecting first available session transport "hla_swd". To override u
se 'transport select <transport>[0m'.
Info : The selected transport took over low-level target control. The results mi
ght differ compared to plain JTAG/SWD
Info : clock speed 2000 kHz
Info : STLINK V2J37S7 (API v2) VID:PID 0483:3748
Info : Target voltage: 3.218361
Info : STM32F439VIT6.cpu: hardware has 6 breakpoints, 4 watchpoints
Info : starting gdb server for STM32F439VIT6.cpu on 3333
Info : Listening on port 3333 for gdb connections
semihosting is enabled
Error: The 'tcl_port' command must be used before 'init'.
But when pressing the second button (debug), it will flash, run the code and start a gdb server all without problems.
The tooling used is:
OpenOCD (version: OpenOCD-20200729-0.10).
Clion (version: 2021.1 but problem persists on previous version) as my IDE.
mingw64 installed via msys.
arm gdb debugger (version 8.3.1)
If you have an idea what the problem could be, please let me know. I can keep using debug without breakpoints but it is annoying as hell that I just can't run.
Somewhere down the line a fix was needed for something.
# Enable semihosting at startup
# Dirty fix for https://youtrack.jetbrains.com/issue/CPP-7103
init
arm semihosting enable
Commenting the init and arm semihosting enable line fixed the problem.
Still weird it would work in debug configuration but not in run configuration. If someone knows why exactly. Please leave an answer in the comments or a post.
Happy coding :)

Gdb can't connect to OpenOCD on stm32

Trying to debug my sample blink_led code on STM32L476 Nucleo-64 board but gdb can't connect to OpenOCD (connection drops almost instantly with error). I've read plenty of posts here and there but none of them helped. Tried adding commands to OpenOCD using -c but no change of behavior.
My code compiles both in Release and Debug config in Eclipse. I can flash the bin file using drag and drop (while the board has built-in STLink add-on) and looks the code runs perfectly on the board (LED blinks).
Cross compiling on Centos7 using the following versions:
Toolchain (gdb): gcc-arm-none-eabi-8-2018-q4-major
OpenOCD: 0.10.0-11-20190118-1134
As using eclipse didn't work I tried the command line,
(I'm not an experienced developer in this environment so I could not find any config file closer to my stm32l476 board than the stm32l4discovery.cfg, please let me know if there might be some issues using it)
./bin/openocd -f scripts/board/stm32l4discovery.cfg -c "init"
It starts,
GNU MCU Eclipse 64-bit Open On-Chip Debugger 0.10.0+dev-00462-gdd1d90111 (2019-01-18-11:37)
Licensed under GNU GPL v2
For bug reports, read
http://openocd.org/doc/doxygen/bugs.html
Info : The selected transport took over low-level target control. The results might differ compared to plain JTAG/SWD
adapter speed: 500 kHz
adapter_nsrst_delay: 100
none separate
srst_only separate srst_nogate srst_open_drain connect_deassert_srst
Info : clock speed 500 kHz
Info : STLINK V2J28M17 (API v2) VID:PID 0483:374B
Info : Target voltage: 3.244386
Info : stm32l4x.cpu: hardware has 6 breakpoints, 4 watchpoints
Info : Listening on port 3333 for gdb connections
Info : Listening on port 6666 for tcl connections
Info : Listening on port 4444 for telnet connections
Then starting GDB:
./arm-none-eabi-gdb ~/eclipse-workspace/test-blink-led/Debug/test-blink-led.elf
then running the following command in gdb:
(gdb) target remote localhost:3333
Remote debugging using localhost:3333
Remote connection closed
As it shows gdb connection drops instantly and OpenOCD prompts the following errors:
Info : accepting 'gdb' connection on tcp/3333
target halted due to debug-request, current mode: Thread
xPSR: 0x01000000 pc: 0x080022e6 msp: 0x20017ff8
Info : device id = 0x10076415
Warn : STM32 flash size failed, probe inaccurate - assuming 1024k flash
Info : flash size = 1024kbytes
Error: auto_probe failed
Error: Connect failed. Consider setting up a gdb-attach event for the target to prepare target for GDB connect, or use 'gdb_memory_map disable'.
Error: attempted 'gdb' connection rejected
Error: jtag status contains invalid mode value - communication failure
Polling target stm32l4x.cpu failed, trying to reexamine
Examination failed, GDB will be halted. Polling again in 100ms
So from those geeks who do it on a similar platform on a daily basis, can anyone help and tell me where am I doing wrong. Does missing any compile-time flag might result in this problem?
I'm scratching my head for a couple of days now so please let me have your hints.
One of the reasons may be that your STLINK firmware seems pretty old (STLINK V2J28M17 as your log shows). I suggest downloading the STSW-LINK007 application to upgrade the firmware. The software is a multiplatform Java application. It works flawlessly in Debian GNU/Linux.
Currently, I use another gdb server texane/stlink for my debugging task with GDB without any problem on some Nucleo and also custom boards. I use target extended-remote command to join the port of the server. Maybe you can try to connect with this command also under OpenOCD.
Try
telnet localhost 4444
it worked for me, while 3333 didn't

Use of target remote localhost:26000

I have been trying to debug my xv6 locally (It is a 32-bit code). But I am working on a 64 bit machine and my gdb is 64 bit. Whenever I type
make gdb
I encountered the following error
+ target remote localhost:26000
.gdbinit:23: Error in sourced command file:
localhost:26000: Connection timed out.
(gdb) target remote localhost:26000
localhost:26000: Connection timed out.
I have a .gdbinit file. I have the following line in it
target remote localhost:26000
It's returning an error at this line. I initially thought it is because of 32-bit architecture of my machine. But if I go to some other directory and type the same message in the gdb console I am receiving the same error. I can't actually understand why it should connect to localhost(i.e what is need of a server here? ) as GDB is a normal GNU debugger right? and what is the reason for its failure in the above case?
Thanks in advance
Your question doesn't make any sense (to me).
As far as I can tell, you are asking: "why does GDB try to connect to localhost:2600".
The answer is: because you asked GDB to do so with the target remote localhost:26000 entry in your ~/.gdbinit.
Update:
I am asking what does that command do?
It instructs GDB to perform remote debugging, and to connect to gdbserver listening on the given port.
In general, you should never put anything into your ~/.gdbinit (or any other initialization file) that you don't understand. Doing so is akin to to taking an unknown object (a gun), putting it to your head, and squeezing a little lever on the side (a trigger).
and executing it giving me the above error so can you please say what is the error? and how to resolve it
Just delete the command (and any other commands that you don't know the meaning of) from ~/.gdbinit.
I've just solved the same problem u described. U can have a look at the GNUmakefile to search the port number your gdb-qemu attempted to connect to. In my GNUmakefile, the port number is
# try to generate a unique GDB port
GDBPORT := $(shell exprid -u% 5000 + 25000)
So when I try to run sudo make gdb-qemu, I got the result as follows:
qemu-system-i386 -drive file=obj/kern/kernel.img,index=0,media=disk,format=raw -serial mon:stdio -gdb tcp::25000 -D qemu.log -S
But when I run another terminal to run sudo make gdb, I got the result:
Type "apropos word" to search for commands related to "word".
+ target remote localhost:26000
The port number didn't match.
So by simply modify the port number in the GNUmakefile to 26000, we can get the ideal results.
Hope it's helpful :)
I had ever met the same problem as you. I had been troubled by this problem for hours.
But this problem is so easy.Please excute the following commands .
make qemu-gdb
make gdb (in another terminal.)
The reason for this problem is that when you execute the make gdb command, you end the execution of the make gdb .So the qemu server connection timed out .
In the picture, we can't alse find the infomation like remote gdb port...
You should change the port in the .gdbint file. It tells GDB which port to detach. I solved this problem in this way.

Remote GDB disconnects whenever I press control + c

I am remote debugging a Stellaris Launchpad. I use OpenOCD to connect to the stellaris and then connect GDB to the server provided by openOCD. I use Open On-Chip Debugger 0.10.0-dev-00002-g79fdeb3 (2015-07-09-23:28). GDB is the one from arm-gcc-none-eabi, the 4_9-2015q1 release.
I invoke openOCD like this:
/usr/local/bin/openocd --file \
/usr/local/share/openocd/scripts/board/ek-lm4f120xl.cfg \
>> openocdLog.txt 2>&1 &
And then GDB like this:
arm-none-eabi-gdb proj//debug/exec -x gdb//gdb.script
gdb/gdb.script contains:
set remotetimeout 10000
target extended-remote :3333
monitor reset halt
load
monitor reset init
The problem is that whenever I hit control+c GDB disconnects. Normally this would halt the remote, but GDB just disconnects:
(gdb) cont
Continuing.
^CError detected on fd 6
Remote communication error. Target disconnected.: Interrupted system call.
(gdb)
OpenOCD has the following things to say, this one while GDB is launching:
Warn : keep_alive() was not invoked in the 1000ms timelimit. GDB alive packet not sent! (1258). Workaround: increase "set remotetimeout" in GDB
Which is weird, considering the gdb/gdb.script file forces remotetimeout to an insanly large number.
And when pressing control+c openOCD says:
Debug: 2602 5089 hla_interface.c:119 hl_interface_quit(): hl_interface_quit
So, how do I resolve this? How can I make GDB halt the remote instead of disconnecting when pressing control+c?
The problem was OpenOCD being too bleeding edge. I had issues with 0.6.1, but version 0.7.0 of OpenOCD works great.

GDB remote protocol: how to analyse packets?

I have:
A proprietary prototype ARM board (Cortex-M3 based) with eCos OS
The board has the programmed RedBoot bootloader
Serial line (RS-232)
GDB debugger for ARM (arm-eabi-gdb)
Host OS is Windows/Cygwin and/or Linux (actually, doesn't matter)
Problem: GDB debugger cannot connect to the target over the serial line.
What I want: is to sniff the packets of the GDB remote protocol in order to undestand whether the GDB stub on the target is alive and operating.
Details: RedBoot has an option to pass the target's control to the built-in GDB stub. I know that the RedBoot is alive, I can connect to it and send it commands over the serial line. The RedBoot manual says that the switch to the GDB stub can be made by typing $ or + symbols (which are actually the prefixes of the GDB remote protocol packets). It seems to work for when I send those symbols the terminal dies. But I'm not sure if the RedBoot was compiled with the GDB stub support (don't ask me why :-)).
Then, when I try to connect to the board with my GDB debugger I get the following picture (on Windows):
(gdb) target remote COM3
Remote debugging using COM3
Ignoring packet error, continuing...
Ignoring packet error, continuing...
Ignoring packet error, continuing...
Ignoring packet error, continuing...
The port is correct, the baudrate as well. Actually the same output I get if I try to do the same with another serial port which is not connected with anything.
What I want to know is does GDB stub send back anything or not?
Intuitively I thought that probably
set verbose on
would help, but GDB manual says it has very a limited effect and my case is beyond it.
May be it is possible to compile GDB debugger with a macro which enables debug logging?
What I want to know is does GDB stub send back anything or not?
Do this (before target remote):
(gdb) set debug remote 1