newlib-cygwin/winsup/cygwin/how-to-debug-cygwin.txt

129 lines
5.5 KiB
Plaintext
Raw Normal View History

Contributed by Egor Duda
2001-09-15 01:43:17 +08:00
2002-01-30 10:44:21 +08:00
So, your favorite program has crashed? And did you say something about
'stackdump'? Or it just prints its output from left to right and
upside-down? Well, you can file an angry bug report and wait until some
of the core developers try to reproduce your problem, try to find what's
the matter with your program and cygwin and fix the bug, if any. But
you can do something better than that. You can debug the problem
yourself, and even if you can't fix it, your analysis may be very
helpful. Here's the (incomplete) howto on cygwin debugging.
2002-07-14 05:13:37 +08:00
1. First things first
The first thing you'll need to do is to build cygwin1.dll and your
2002-01-30 10:44:21 +08:00
crashed application from sources. To debug them you'll need debug
information, which is normally stripped from executables. You probably
also want to build a version of the dll with more debugging capabilities
by reconfiguring your build directory, specifying the --enable-debugging
option to configure.
2001-09-15 01:43:17 +08:00
2002-07-14 05:13:37 +08:00
2. Creating a known-working cygwin debugging environment
2002-01-30 10:44:21 +08:00
- create a separate directory, say, c:\cygdeb, and put known-working
cygwin1.dll and gdb.exe in it.
- create a wrapper c:\cygdeb\debug_wrapper.cmd:
2001-09-15 01:43:17 +08:00
========= debug_wrapper.cmd =========
2001-10-10 02:12:51 +08:00
rem setting CYGWIN_TESTING environment variable makes cygwin application
2001-09-15 01:43:17 +08:00
rem not to interfere with other already running cygwin applications.
set CYGWIN_TESTING=1
c:\cygdeb\gdb.exe -nw %1 %2
===================================
2002-07-14 05:13:37 +08:00
3. Using cygwin's JIT debugging facility
2002-01-30 10:44:21 +08:00
add 'error_start=c:\cygdeb\debug_wrapper.cmd' to CYGWIN environment
variable. When some application encounters critical error, cygwin will stop
it and execute debug_wrapper.cmd, which will run gdb and make it to attach to
the crashed application.
2001-09-15 01:43:17 +08:00
2002-07-14 05:13:37 +08:00
4. Strace
2002-01-30 10:44:21 +08:00
You can run your program under 'strace' utility, described if user's manual.
If you know where the problem approximately is, you can add a bunch of
additional debug_printf()s in the source code and see what they print in
strace log. There's one common problem with this method, that some bugs
may mysteriously disappear once the program is run under strace. Then the
bug is likely a race condition. strace has two useful options to deal with
such situation: -b enables buffering of output and reduces additional
timeouts introduced by strace, and -m option allows you to mask certain
classes of *_printf() functions, reducing timeouts even more.
Note that strace does not use the cygwin DLL and so any process that it
starts does not inherit a cygwin environment. It is equivalent to starting
a program from the command prompt.
2001-09-15 01:43:17 +08:00
2002-07-14 05:13:37 +08:00
5. Problems at early startup
2002-01-30 10:44:21 +08:00
Sometimes, something crashes at the very early stages of application
initialization, when JIT debugging facility is not yet active. Ok, there's
another environment variable that may help. Create program_wrapper.cmd:
2001-09-15 01:43:17 +08:00
========= program_wrapper.cmd =========
2001-10-10 02:12:51 +08:00
rem setting CYGWIN_SLEEP environment variable makes cygwin application
2001-09-15 01:43:17 +08:00
rem to sleep for x milliseconds at startup
set CYGWIN_SLEEP=20000
c:\some\path\bad_program.exe some parameters
===================================
2001-12-10 11:35:54 +08:00
2002-01-30 10:44:21 +08:00
Now, run program_wrapper.cmd. It should print running program pid.
After starting program_wrapper.cmd you've got 20 seconds to open another
window, cd to c:\cygdeb in it, run gdb there and in gdb prompt type
2001-09-15 01:43:17 +08:00
2002-01-30 10:44:21 +08:00
(gdb) attach <pid>
2001-09-15 01:43:17 +08:00
2002-01-30 10:44:21 +08:00
where <pid> is the pid that program_wrapper.cmd have printed.
After that you can normally step through the code in cygwin1.dll and
bad_program.exe
2001-09-15 01:43:17 +08:00
2002-07-14 05:13:37 +08:00
6. More problems at early startup
You can also set a CYGWIN_DEBUG variable to force the debugger to pop up
only when a certain program is run:
2005-04-25 13:03:53 +08:00
set CYGWIN_DEBUG=cat.exe:gdb.exe
This will force gdb.exe to start when the program name contains the string
2005-04-25 13:03:53 +08:00
"cat.exe". The ':gdb.exe' isn't really needed, since it is the default.
It is just there to show how you can specify a program to run when the
2002-07-14 05:15:10 +08:00
program starts. You can optionally set a breakpoint on "break_here"
once the debugger pops up and then you can single step through the
initialization process.
Note that it bears repeating that both of the above options are *only*
available when configuring cygwin with --enable-debugging.
2002-07-14 05:13:37 +08:00
7. Heap corruption
2002-01-30 10:44:21 +08:00
If your program crashes at malloc() or free() or when it references some
malloc()'ed memory, it looks like heap corruption. You can configure and
build special version of cygwin1.dll which includes heap sanity checking.
To do it, just add --enable-malloc-debugging option to configure. Be warned,
however, that this version of dll is _very_ slow (10-100 times slower than
normal), so use it only when absolutely necessary.
2001-11-05 11:16:58 +08:00
2002-07-14 05:13:37 +08:00
8. Program dies when running under strace
2002-01-30 10:44:21 +08:00
If your program crashes when you run it using strace but runs ok (or has a
different problem) otherwise, then there may be a problem in one of the
strace *_printf statements. Usually this is caused by a change in arguments
resulting in a %s being used with something other than a pointer to a
string.
To debug this scenario, do something like this:
bash$ gdb -nw yourapp.exe
(gdb) dll cygwin1
(gdb) l dll_crt0_1
2005-05-07 05:09:35 +08:00
(gdb) b <<first line in the function>>
2002-01-30 10:44:21 +08:00
(gdb) run
(gdb) set strace._active=1
2002-01-30 10:44:21 +08:00
(gdb) continue
The program will then run in "strace mode", calling each strace *_printf,
just like it does when run under the strace program. Eventually, the
program will crash, probably in small_printf. At that point, a 'bt'
command should show you the offending call to strace_printf with the
improper format string.