[Date Prev][Date Next]   [Thread Prev][Thread Next]   [Thread Index] [Date Index] [Author Index]

Re: Problem in rollback of failed transactions...



On Tue, 3 Jun 2003, James Olin Oden wrote:
<snip>
> 	1) In my function that finds the repackaged package at the
> 	   very end I free the IDTX (using IDTXfree()).  When I do 
> 	   this I get a segfault.  If I don't do that (which is got
> 	   to be a monstrous memory leak) it does not segfault.
> 	2) If don't free the IDTX, and the code can continue, then when
> 	   the rollback transaction is ran it says that repackaged 
> 	   package is already installed.
> 
> On the first point, I was wondering if there were any pointers anyone
> could give me such that I could keep the segfault from occuring?
> I am using headerLink() liberally BTW.
>
I think I got this one.  The trick is you never do:

	another_header = some_header;

you always do:

	another_header = headerLink(header);

And the header link segfault madness goes away (-;
 
> On the second point, since I am only adding transaction elements 
> after a successful run through the psm, I figured that the header 
> from the old package would be gone?  What do I need to do to deal 
> with this issue?  Any ideas?
>
I am still trying to figure this one out.  I have scaned through the psm
yet again (I might get it eventually) and it sure looks like that the 
appending of the new header and removing of the old header happens in 
PSM_POST but it sure does not look like that is happening when I run the
rollback transaction.  Do I need to somehow need to commit the failed 
transaction before I run the rollback transaction; or close the failed
transactions db?

Simply clueless...james
> Thanks...james
> 
> 
> _______________________________________________
> Rpm-list mailing list
> Rpm-list@redhat.com
> https://www.redhat.com/mailman/listinfo/rpm-list
> 




[Date Prev][Date Next]   [Thread Prev][Thread Next]   [Thread Index] [Date Index] [Author Index] []