Linux kernel mirror (for testing) git.kernel.org/pub/scm/linux/kernel/git/torvalds/linux.git
kernel os linux
1
fork

Configure Feed

Select the types of activity you want to include in your feed.

at v2.6.23-rc9 179 lines 7.8 kB view raw
1 2This file describes the configuration and behavior of KGDB for the SH 3kernel. Based on a description from Henry Bell <henry.bell@st.com>, it 4has been modified to account for quirks in the current implementation. 5 6Version 7======= 8 9This version of KGDB was written for 2.4.xx kernels for the SH architecture. 10Further documentation is available from the linux-sh project website. 11 12 13Debugging Setup: Host 14====================== 15 16The two machines will be connected together via a serial line - this 17should be a null modem cable i.e. with a twist. 18 19On your DEVELOPMENT machine, go to your kernel source directory and 20build the kernel, enabling KGDB support in the "kernel hacking" section. 21This includes the KGDB code, and also makes the kernel be compiled with 22the "-g" option set -- necessary for debugging. 23 24To install this new kernel, use the following installation procedure. 25 26Decide on which tty port you want the machines to communicate, then 27cable them up back-to-back using the null modem. On the DEVELOPMENT 28machine, you may wish to create an initialization file called .gdbinit 29(in the kernel source directory or in your home directory) to execute 30commonly-used commands at startup. 31 32A minimal .gdbinit might look like this: 33 34 file vmlinux 35 set remotebaud 115200 36 target remote /dev/ttyS0 37 38Change the "target" definition so that it specifies the tty port that 39you intend to use. Change the "remotebaud" definition to match the 40data rate that you are going to use for the com line (115200 is the 41default). 42 43Debugging Setup: Target 44======================== 45 46By default, the KGDB stub will communicate with the host GDB using 47ttySC1 at 115200 baud, 8 databits, no parity; these defaults can be 48changed in the kernel configuration. As the kernel starts up, KGDB will 49initialize so that breakpoints, kernel segfaults, and so forth will 50generally enter the debugger. 51 52This behavior can be modified by including the "kgdb" option in the 53kernel command line; this option has the general form: 54 55 kgdb=<ttyspec>,<action> 56 57The <ttyspec> indicates the port to use, and can optionally specify 58baud, parity and databits -- e.g. "ttySC0,9600N8" or "ttySC1,19200". 59 60The <action> can be "halt" or "disabled". The "halt" action enters the 61debugger via a breakpoint as soon as kgdb is initialized; the "disabled" 62action causes kgdb to ignore kernel segfaults and such until explicitly 63entered by a breakpoint in the code or by external action (sysrq or NMI). 64 65(Both <ttyspec> and <action> can appear alone, w/o the separating comma.) 66 67For example, if you wish to debug early in kernel startup code, you 68might specify the halt option: 69 70 kgdb=halt 71 72Boot the TARGET machine, which will appear to hang. 73 74On your DEVELOPMENT machine, cd to the source directory and run the gdb 75program. (This is likely to be a cross GDB which runs on your host but 76is built for an SH target.) If everything is working correctly you 77should see gdb print out a few lines indicating that a breakpoint has 78been taken. It will actually show a line of code in the target kernel 79inside the gdbstub activation code. 80 81NOTE: BE SURE TO TERMINATE OR SUSPEND any other host application which 82may be using the same serial port (for example, a terminal emulator you 83have been using to connect to the target boot code.) Otherwise, data 84from the target may not all get to GDB! 85 86You can now use whatever gdb commands you like to set breakpoints. 87Enter "continue" to start your target machine executing again. At this 88point the target system will run at full speed until it encounters 89your breakpoint or gets a segment violation in the kernel, or whatever. 90 91Serial Ports: KGDB, Console 92============================ 93 94This version of KGDB may not gracefully handle conflict with other 95drivers in the kernel using the same port. If KGDB is configured on the 96same port (and with the same parameters) as the kernel console, or if 97CONFIG_SH_KGDB_CONSOLE is configured, things should be fine (though in 98some cases console messages may appear twice through GDB). But if the 99KGDB port is not the kernel console and used by another serial driver 100which assumes different serial parameters (e.g. baud rate) KGDB may not 101recover. 102 103Also, when KGDB is entered via sysrq-g (requires CONFIG_KGDB_SYSRQ) and 104the kgdb port uses the same port as the console, detaching GDB will not 105restore the console to working order without the port being re-opened. 106 107Another serious consequence of this is that GDB currently CANNOT break 108into KGDB externally (e.g. via ^C or <BREAK>); unless a breakpoint or 109error is encountered, the only way to enter KGDB after the initial halt 110(see above) is via NMI (CONFIG_KGDB_NMI) or sysrq-g (CONFIG_KGDB_SYSRQ). 111 112Code is included for the basic Hitachi Solution Engine boards to allow 113the use of ttyS0 for KGDB if desired; this is less robust, but may be 114useful in some cases. (This cannot be selected using the config file, 115but only through the kernel command line, e.g. "kgdb=ttyS0", though the 116configured defaults for baud rate etc. still apply if not overridden.) 117 118If gdbstub Does Not Work 119======================== 120 121If it doesn't work, you will have to troubleshoot it. Do the easy 122things first like double checking your cabling and data rates. You 123might try some non-kernel based programs to see if the back-to-back 124connection works properly. Just something simple like cat /etc/hosts 125/dev/ttyS0 on one machine and cat /dev/ttyS0 on the other will tell you 126if you can send data from one machine to the other. There is no point 127in tearing out your hair in the kernel if the line doesn't work. 128 129If you need to debug the GDB/KGDB communication itself, the gdb commands 130"set debug remote 1" and "set debug serial 1" may be useful, but be 131warned: they produce a lot of output. 132 133Threads 134======= 135 136Each process in a target machine is seen as a gdb thread. gdb thread related 137commands (info threads, thread n) can be used. CONFIG_KGDB_THREAD must 138be defined for this to work. 139 140In this version, kgdb reports PID_MAX (32768) as the process ID for the 141idle process (pid 0), since GDB does not accept 0 as an ID. 142 143Detaching (exiting KGDB) 144========================= 145 146There are two ways to resume full-speed target execution: "continue" and 147"detach". With "continue", GDB inserts any specified breakpoints in the 148target code and resumes execution; the target is still in "gdb mode". 149If a breakpoint or other debug event (e.g. NMI) happens, the target 150halts and communicates with GDB again, which is waiting for it. 151 152With "detach", GDB does *not* insert any breakpoints; target execution 153is resumed and GDB stops communicating (does not wait for the target). 154In this case, the target is no longer in "gdb mode" -- for example, 155console messages no longer get sent separately to the KGDB port, or 156encapsulated for GDB. If a debug event (e.g. NMI) occurs, the target 157will re-enter "gdb mode" and will display this fact on the console; you 158must give a new "target remote" command to gdb. 159 160NOTE: TO AVOID LOSSING CONSOLE MESSAGES IN CASE THE KERNEL CONSOLE AND 161KGDB USING THE SAME PORT, THE TARGET WAITS FOR ANY INPUT CHARACTER ON 162THE KGDB PORT AFTER A DETACH COMMAND. For example, after the detach you 163could start a terminal emulator on the same host port and enter a <cr>; 164however, this program must then be terminated or suspended in order to 165use GBD again if KGDB is re-entered. 166 167 168Acknowledgements 169================ 170 171This code was mostly generated by Henry Bell <henry.bell@st.com>; 172largely from KGDB by Amit S. Kale <akale@veritas.com> - extracts from 173code by Glenn Engel, Jim Kingdon, David Grothe <dave@gcom.com>, Tigran 174Aivazian <tigran@sco.com>, William Gatliff <bgat@open-widgets.com>, Ben 175Lee, Steve Chamberlain and Benoit Miller <fulg@iname.com> are also 176included. 177 178Jeremy Siegel 179<jsiegel@mvista.com>