[Gc] bug report format
hans.boehm at hp.com
Tue Jan 31 10:29:22 PST 2006
There is no fixed format. However, I'm generally overcommitted, so
things that can be quickly dealt with are far more likely to get prompt
attention. I expect the same is true for others who tend to contribute
patches. My preferences would be, in order:
1) A patch that includes an added regression test in gctest. (Wishful
thinking, I know.)
2) A patch.
3) A precise description of why the code breaks.
4) A small test case that reproduces the problem on either Linux/X86 or
Linux/IA64. (Or possibly some other machine I can get access to. More
common platforms tend to generate more interest.)
5) A gdb log (or equivalent) that suggests something about what went
wrong. This is likely to require multiple iterations on the mailing
list. (Look at
http://www.hpl.hp.com/personal/Hans_Boehm/gc/debugging.html for some
6) A rough description of what went wrong, together with any relevant
data. For premature collection problems, whatever gdb output you can
gather, including especially any information about which objects on the
chain of references to the collected objects were marked, is very
helpful. For performance issues, generally GC_PRINT_STATS and/or
GC_DUMP_REGULARLY output is useful. (The latter may be too big to post
In any case, please post to this list.
In your case, it would be good to know if GC_all_interior_pointers was
You might also want to look at the "Strange Issues" thread on this list,
which sounds similar.
> -----Original Message-----
> From: gc-bounces at napali.hpl.hp.com
> [mailto:gc-bounces at napali.hpl.hp.com] On Behalf Of Paul Biggar
> Sent: Tuesday, January 31, 2006 3:29 AM
> To: gc at napali.hpl.hp.com
> Subject: [Gc] bug report format
> We'd like to submit a bug report regarding segfaults in our
> single-threaded C++ program (http://www.phpcompiler.org, if
> you're interested). The bug is probably related to dynamic
> casting and multiple inheritence, since we do a lot of this.
> However, we do not know the intimate details of the garbage
> collector, and so our initial efforts at debugging havent
> gone very far.
> Could somebody point us to a format for submitting bug
> reports to you? Do you need stack traces or core dumps, or
> our code, or do you only accept patches?
> Thanks in advance,
> Paul Biggar
> paul.biggar at gmail.com
> Gc mailing list
> Gc at linux.hpl.hp.com
More information about the Gc