"David D. Pagan" wrote: > Roland Mainz wrote: > > Roland Mainz wrote: > >> Alexander.Gorshenev at Sun.COM wrote: > >>> On Wed, 19 Mar 2008, Roland Mainz wrote: > >>>> Keith M Wesolowski wrote: > >>>>> On Sat, Mar 15, 2008 at 04:47:25PM +0100, Roland Mainz wrote: [snip] > > ... now I have to find the Bugster CR number and try to escalate it... > > An extension to the -xcheck flag, [no%]init_local, was included in Sun > Studio 12. See my blog entry: > > http://blogs.sun.com/dpagan/entry/sun_studio_12_simplifying_the > > Also, in an upcoming express release, this flag has been extended in a > couple of ways. Check out this blog entry: > > http://blogs.sun.com/dpagan/entry/update_to_a_useful_feature > > I must have missed the specifics of your request, so I can't comment on > how well the new -xcheck functionality addresses your needs.
What we need is something which allows us to specify a fixed byte sequence (in variable width, e.g. 0xFF will repeatedly write 0xFFFFFFFF... into the memory while 0xDEADBEEF00 will write 0xDEADBEEF00DEADBEEF00DEADBEEF00... into the memory). The idea is to make the related issues caused by uninitalised variables reproducible, test various combinations of values and have a fixed value which can be spotted in core dumps. > However, it > doesn't currently support user supplied initialization values (as > mentioned in the email below). Check out the links above > to see if this feature is of any help. Any enhancements/changes > requested (including the one you already submitted) will always be > considered for an upcoming release. Umpf... it's needed in Studio 12 (the upcoming release used for OS/Net). > If you have any other comments/questions, let me know ... See above... ---- Bye, Roland -- __ . . __ (o.\ \/ /.o) roland.mainz at nrubsig.org \__\/\/__/ MPEG specialist, C&&JAVA&&Sun&&Unix programmer /O /==\ O\ TEL <currently fluctuating> (;O/ \/ \O;)
