cvs commit: src/sys/alpha/alpha support.s src/sys/i386/i386
swtch.s src/sys/kern kern_shutdown.c src/sys/sys systm.h
bde at zeta.org.au
Mon Jan 19 23:21:24 PST 2004
On Mon, 19 Jan 2004, Jeff Roberson wrote:
> On Mon, 19 Jan 2004, Poul-Henning Kamp wrote:
> > phk 2004/01/19 13:27:11 PST
> > FreeBSD src repository
> > Modified files:
> > sys/alpha/alpha support.s
> > sys/i386/i386 swtch.s
> > sys/kern kern_shutdown.c
> > sys/sys systm.h
> > Log:
> > Add linenumber and source filename to panic(9) output.
> > Ideally a traceback should be printed too, any takers ?
> So, how much effect on datasize did this actually have?
Obviously too much to be good :-(. It probably bloats text size more
than data size, since the strings for file names are shared.
There are 2885 references to panic and 1458 to KASSERT, so a simople
estimate gives 4000 * 17 = 68K of code in LINT for the text space
bloat on i386's.
Other bugs in this commit include printing the file and line number
twice for panics that already print them once. All 4000+ calls need
to be examined to find these bugs. Mutex debugging has panics like:
panic("mutex %s not owned at %s:%d",
m->mtx_object.lo_name, file, line);
where `file' and `line' for the caller are relevant but __FILE__
and __LINE__ for the mutex function are just those of the messenger.
More information about the cvs-all