Re: [Evolution] SMP Safe?

[Not Zed <notzed ximian com>]

Oh debian.  Send them the bug reports then ... they make changes to our
code and we can't support it.

Well, seriously ... i'm thinking of getting an smp machine sometime
soon, so perhaps i can have a better go at chasing these issues up.

AFAIK there shouldn't be too many issues (many people use it
successfully), but i'm not entirely confident of the code, wrt
compilation optimisations and whatnot (possibly needs a few volatile's
thrown around in certain places).

On Mon, 2003-03-03 at 19:59, Mika Liljeberg wrote:
> On Mon, 2003-03-03 at 05:03, Not Zed wrote:
> > did you submit any bug reports?
>
> Not recently. The problem is that the debian/sid Evolution package
> doesn't have debug syms, so the backtraces aren't very useful. I
> submitted a couple of those into bugzilla earlier and they gor rejected
> out of hand.
>
> Compiling for debug takes forever on my machine and I didn't bother to
> do that last time when evolution got pulled from the debian repository.
> Last time (couple of months ago with evo-1.2.1 I think) I had a real go
> at tracking down the crashes I ended up with this:
>
> "I had another crash in a completely different place (syncing the index
> file of my Sent folder to disk). One pointer on the block list had a
> next pointer with the value 0x200. And here we had a pointer with the
> value 0x100. Looks very suspicious to me. I don't suppose these numbers
> mean anything to you? Could be a flags field, of some sort, I guess."
>
> So, I was seeing similarily garbled pointers in different lists and
> crashes in places that traversed those lists. Unfortunately, the above
> didn't elicit any comments on the mailing list and I don't know enough
> about the code to track down the cause. If you want to give me a hand
> I'll give it another go. Recompiling now.

Hmm, rings a bell, i thought i replied to that, but probably without
anything meaningful.