[Date Prev][Date Next] [Chronological] [Thread] [Top]

Re: deadlock in back-ldbm/modrdn.c (ITS#951)

On Mon, Jan 08, 2001 at 01:00:29PM -0800, Kurt D. Zeilenga wrote:
> >The old and the new parent is the same, and it looks like it tries to write
> >lock
> >the parent twice (both the old and the new which are the same).
> We should catch this and either:
>         1) ignore the newSuperior field or
>         2) return an error as the superior is not new.

I commited to head a fix that does 1. Please see if it looks reasonable,
seems to work okay.