Skip to contentSkip to conversation

Usage of kgdb in v2.6.26

DISCUSSION ZC Miao 2 participants 4 messages Jul 16, 2008 – Jul 17, 2008

lore
Share & tools
lore

ZC Miao

author
#0 lore raw

I tried to use the kgdb from v2.6.26, but met few problems. I use :
1. gdb 6.8 (gdb-6.8-11.fc9.i386)
2. kernel v2.6.26 source code
3. Qemu 0.9.1 (qemu-0.9.1-5.fc9.i386)
4. Host is : 2.6.25.10-86.fc9.i686
5. Cpu : Intel(R) Core(TM)2 Duo CPU T5450 @ 1.66GHz

* Firstly I compile kernel 2.6.26 with kgdb support :
CONFIG_KGDB=y
CONFIG_KGDB_SERIAL_CONSOLE=y

* Then I use qemu to boot my kernel along with a rootfs that I created, with a qemu option that makes a local tcp port a listening service as a serial port for the emulated system. The kernel arguments I gave is :
nfsroot=192.168.200.1:/path/to/rootfs,udp,nolock ip=dhcp root=/dev/nfs ro single kgdboc=ttyS0,115200 kgdbwait

* After the kernel loaded and boot for a while, the gdb receives a signal and hangs the kernel :
kgdb_register_io_module (new_kgdb_io_ops=<value optimized out>)

    at /data/home/hellwolf/mydoc/prog/linux/linux-2.6/kernel/kgdb.c:1674

(gdb)

Question 1 :
* If I type "b sys_read", and then "c", the kernel continues until I get :

---------------------------------------------------------------
[New Thread 1]
Cannot remove breakpoints because program is no longer writable. It might be running in another process.
Further execution is probably impossible.

Breakpoint 1, sys_read (fd=3, buf=0xbfdb31dc "", count=512)

at /data/home/hellwolf/mydoc/prog/linux/linux-2.6/fs/read_write.c:360 (gdb)
---------------------------------------------------------------

Why "program is no longer writable" happens here?

Question 2 :
* If I directly type "c" when kgdbwait triggered. And use Sys-Rq+G to interrupt the kernel, I will get :
---------------------------------------------------------------
[New Thread -1]

Program received signal SIGTRAP, Trace/breakpoint trap. [Switching to Thread -1]
sysrq_handle_gdb (key=103, tty=0xc78f6400)

    at /data/home/hellwolf/mydoc/prog/linux/linux-2.6/kernel/kgdb.c:167

(gdb)
---------------------------------------------------------------
Here if I type "c sys_read" and "c", it never actually makes any break points, I will never receive a break from gdb.

I don't know if I made anything wrong, or I missed any important thing, so I post here and would like to get some help from you.

Thanks.

--
ZC Miao ([email protected])
Blog http://hellwolf.cublog.cn

gpg --keyserver pgp.mit.edu --recv-key 0x6B174C6F

Most people deserve each other.
-- Shirley

#1 lore raw

Do you have CONFIG_DEBUG_RODATA=y in your kernel config?

There is a presently a known problem where breakpoints will silently fail to plant when this is turned on.

Jason.

Quoted 80 lines · ZC Miao

I tried to use the kgdb from v2.6.26, but met few problems. I use :
1. gdb 6.8 (gdb-6.8-11.fc9.i386)
2. kernel v2.6.26 source code
3. Qemu 0.9.1 (qemu-0.9.1-5.fc9.i386)
4. Host is : 2.6.25.10-86.fc9.i686
5. Cpu : Intel(R) Core(TM)2 Duo CPU T5450 @ 1.66GHz

* Firstly I compile kernel 2.6.26 with kgdb support :
CONFIG_KGDB=y
CONFIG_KGDB_SERIAL_CONSOLE=y

* Then I use qemu to boot my kernel along with a rootfs that I created, with a qemu option that makes a local tcp port a listening service as a serial port for the emulated system. The kernel arguments I gave is :
nfsroot=192.168.200.1:/path/to/rootfs,udp,nolock ip=dhcp root=/dev/nfs ro single kgdboc=ttyS0,115200 kgdbwait

* After the kernel loaded and boot for a while, the gdb receives a signal and hangs the kernel :
kgdb_register_io_module (new_kgdb_io_ops=<value optimized out>)

    at /data/home/hellwolf/mydoc/prog/linux/linux-2.6/kernel/kgdb.c:1674

(gdb)

Question 1 :
* If I type "b sys_read", and then "c", the kernel continues until I get :

---------------------------------------------------------------
[New Thread 1]
Cannot remove breakpoints because program is no longer writable. It might be running in another process.
Further execution is probably impossible.

Breakpoint 1, sys_read (fd=3, buf=0xbfdb31dc "", count=512)

at /data/home/hellwolf/mydoc/prog/linux/linux-2.6/fs/read_write.c:360 (gdb)
---------------------------------------------------------------

Why "program is no longer writable" happens here?

Question 2 :
* If I directly type "c" when kgdbwait triggered. And use Sys-Rq+G to interrupt the kernel, I will get :
---------------------------------------------------------------
[New Thread -1]

Program received signal SIGTRAP, Trace/breakpoint trap. [Switching to Thread -1]
sysrq_handle_gdb (key=103, tty=0xc78f6400)

    at /data/home/hellwolf/mydoc/prog/linux/linux-2.6/kernel/kgdb.c:167

(gdb)
---------------------------------------------------------------
Here if I type "c sys_read" and "c", it never actually makes any break points, I will never receive a break from gdb.

I don't know if I made anything wrong, or I missed any important thing, so I post here and would like to get some help from you.

Thanks.

--
ZC Miao ([email protected])
Blog http://hellwolf.cublog.cn

gpg --keyserver pgp.mit.edu --recv-key 0x6B174C6F

Most people deserve each other.
-- Shirley

--
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in the body of a message to [email protected]
More majordomo info at http://vger.kernel.org/majordomo-info.html
Please read the FAQ at http://www.tux.org/lkml/

author
#2 lore raw
$ cat ../../output/x86.kerneldev/.config| grep CONFIG_DEBUG_RODATA
CONFIG_DEBUG_RODATA=y

Yes, conigured.

Quoted 90 lines · Jason Wessel

Do you have CONFIG_DEBUG_RODATA=y in your kernel config?

There is a presently a known problem where breakpoints will silently fail to plant when this is turned on.

Jason.

Quoted 80 lines · ZC Miao

I tried to use the kgdb from v2.6.26, but met few problems. I use :
1. gdb 6.8 (gdb-6.8-11.fc9.i386)
2. kernel v2.6.26 source code
3. Qemu 0.9.1 (qemu-0.9.1-5.fc9.i386)
4. Host is : 2.6.25.10-86.fc9.i686
5. Cpu : Intel(R) Core(TM)2 Duo CPU T5450 @ 1.66GHz

* Firstly I compile kernel 2.6.26 with kgdb support :
CONFIG_KGDB=y
CONFIG_KGDB_SERIAL_CONSOLE=y

* Then I use qemu to boot my kernel along with a rootfs that I created, with a qemu option that makes a local tcp port a listening service as a serial port for the emulated system. The kernel arguments I gave is :
nfsroot=192.168.200.1:/path/to/rootfs,udp,nolock ip=dhcp root=/dev/nfs ro single kgdboc=ttyS0,115200 kgdbwait

* After the kernel loaded and boot for a while, the gdb receives a signal and hangs the kernel :
kgdb_register_io_module (new_kgdb_io_ops=<value optimized out>)

    at /data/home/hellwolf/mydoc/prog/linux/linux-2.6/kernel/kgdb.c:1674

(gdb)

Question 1 :
* If I type "b sys_read", and then "c", the kernel continues until I get :

---------------------------------------------------------------
[New Thread 1]
Cannot remove breakpoints because program is no longer writable. It might be running in another process.
Further execution is probably impossible.

Breakpoint 1, sys_read (fd=3, buf=0xbfdb31dc "", count=512)

at /data/home/hellwolf/mydoc/prog/linux/linux-2.6/fs/read_write.c:360 (gdb)
---------------------------------------------------------------

Why "program is no longer writable" happens here?

Question 2 :
* If I directly type "c" when kgdbwait triggered. And use Sys-Rq+G to interrupt the kernel, I will get :
---------------------------------------------------------------
[New Thread -1]

Program received signal SIGTRAP, Trace/breakpoint trap. [Switching to Thread -1]
sysrq_handle_gdb (key=103, tty=0xc78f6400)

    at /data/home/hellwolf/mydoc/prog/linux/linux-2.6/kernel/kgdb.c:167

(gdb)
---------------------------------------------------------------
Here if I type "c sys_read" and "c", it never actually makes any break points, I will never receive a break from gdb.

I don't know if I made anything wrong, or I missed any important thing, so I post here and would like to get some help from you.

Thanks.

--
ZC Miao ([email protected])
Blog http://hellwolf.cublog.cn

gpg --keyserver pgp.mit.edu --recv-key 0x6B174C6F

Most people deserve each other.
-- Shirley

--
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in the body of a message to [email protected]
More majordomo info at http://vger.kernel.org/majordomo-info.html
Please read the FAQ at http://www.tux.org/lkml/

--
ZC Miao ([email protected])
Blog http://hellwolf.cublog.cn

gpg --keyserver pgp.mit.edu --recv-key 0x6B174C6F

Regarding astral projection, Woody Allen once wrote, "This is not a bad way to travel, although there is usually a half-hour wait for luggage."

author
#3 lore raw
In reply to Jason Wessel

Wow, that's a magic! It works now!

I think we should document it in the Documentation of kgdb.

Thanks!

Quoted 90 lines · Jason Wessel

Do you have CONFIG_DEBUG_RODATA=y in your kernel config?

There is a presently a known problem where breakpoints will silently fail to plant when this is turned on.

Jason.

Quoted 80 lines · ZC Miao

I tried to use the kgdb from v2.6.26, but met few problems. I use :
1. gdb 6.8 (gdb-6.8-11.fc9.i386)
2. kernel v2.6.26 source code
3. Qemu 0.9.1 (qemu-0.9.1-5.fc9.i386)
4. Host is : 2.6.25.10-86.fc9.i686
5. Cpu : Intel(R) Core(TM)2 Duo CPU T5450 @ 1.66GHz

* Firstly I compile kernel 2.6.26 with kgdb support :
CONFIG_KGDB=y
CONFIG_KGDB_SERIAL_CONSOLE=y

* Then I use qemu to boot my kernel along with a rootfs that I created, with a qemu option that makes a local tcp port a listening service as a serial port for the emulated system. The kernel arguments I gave is :
nfsroot=192.168.200.1:/path/to/rootfs,udp,nolock ip=dhcp root=/dev/nfs ro single kgdboc=ttyS0,115200 kgdbwait

* After the kernel loaded and boot for a while, the gdb receives a signal and hangs the kernel :
kgdb_register_io_module (new_kgdb_io_ops=<value optimized out>)

    at /data/home/hellwolf/mydoc/prog/linux/linux-2.6/kernel/kgdb.c:1674

(gdb)

Question 1 :
* If I type "b sys_read", and then "c", the kernel continues until I get :

---------------------------------------------------------------
[New Thread 1]
Cannot remove breakpoints because program is no longer writable. It might be running in another process.
Further execution is probably impossible.

Breakpoint 1, sys_read (fd=3, buf=0xbfdb31dc "", count=512)

at /data/home/hellwolf/mydoc/prog/linux/linux-2.6/fs/read_write.c:360 (gdb)
---------------------------------------------------------------

Why "program is no longer writable" happens here?

Question 2 :
* If I directly type "c" when kgdbwait triggered. And use Sys-Rq+G to interrupt the kernel, I will get :
---------------------------------------------------------------
[New Thread -1]

Program received signal SIGTRAP, Trace/breakpoint trap. [Switching to Thread -1]
sysrq_handle_gdb (key=103, tty=0xc78f6400)

    at /data/home/hellwolf/mydoc/prog/linux/linux-2.6/kernel/kgdb.c:167

(gdb)
---------------------------------------------------------------
Here if I type "c sys_read" and "c", it never actually makes any break points, I will never receive a break from gdb.

I don't know if I made anything wrong, or I missed any important thing, so I post here and would like to get some help from you.

Thanks.

--
ZC Miao ([email protected])
Blog http://hellwolf.cublog.cn

gpg --keyserver pgp.mit.edu --recv-key 0x6B174C6F

Most people deserve each other.
-- Shirley

--
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in the body of a message to [email protected]
More majordomo info at http://vger.kernel.org/majordomo-info.html
Please read the FAQ at http://www.tux.org/lkml/

--
ZC Miao ([email protected])
Blog http://hellwolf.cublog.cn

gpg --keyserver pgp.mit.edu --recv-key 0x6B174C6F

The human mind ordinarily operates at only ten percent of its capacity -- the rest is overhead for the operating system.