Sterling:
I hope the problem is fixed in version 0.44 of Sablotron which is
available for download now. Tell us in case it persists please.
Thanks,
Tom
Sterling Hughes wrote:
>
> Hi guys,
> Here's the message that I got from the person regarding the crash in
> Sablotron...
>
> He can be contacted at: [EMAIL PROTECTED]
>
> Sterling
>
> > Hello,
> >
> > I cut down the problem and I got to the following conclusion:
> >
> > >From datastr.h (in Salbot--0.43\Sablot\engine)
> >
> > 498 template <class T>
> > 499 void List<T>::append(T what)
> > 500 {
> > 501 if (nItems >= blocksize)
> > 502 {
> > 503 if (block)
> > 504 grow();
> > 505 else
> > 506 {
> > 507 blocksize = origBlocksize;
> > 508 block = (T*) claimMemory(blocksize * sizeof(T));
> > 509 // FIXME: asserting memory request ok
> > printf ("Bsize: %04X\n", blocksize);
> > printf ("sizeof(T): %lu\n", sizeof(T));
> > 510 assert(block);
> > 511 }
> > 512 }
> > 513 block[nItems++] = what;
> > 514 };
> >
> > When I get the error mentioned earlier, The debug statements I added print:
> > Bsize: 4000000
> >
> > I think that's way to much memory to claim.
> >
> > I don't think you can/must resolve this problem, but please pass this through
> to
> > to the Sablotron guys. (Or you can let them contact me directly, if that's
> simpler
> > for you).
> >
>
> And here is the relevant part of another message relating to this problem:
>
> -- Start --
> Hello Sterling,
>
> I did have problems with it, but you recently changed the something in the
> extension
>
> and it is now much better. But I still get the following strange error in my
> error_log (Apache) part of the time:
>
> httpd: datastr.h:510: void List<HashItem *>::append<HashItem *>(class HashItem
> *):
> Assertion `block' failed.
> [Wed Sep 6 14:28:23 2000] [notice] child pid 4016 exit signal Aborted (6)
>
> --- End --
>
> The change I made was to simply reuse the current processor when there was
> already an instance of the Sablotron processor.
