wisner@altdorf.AI.MIT.EDU (Bill Wisner) (09/07/88)
I recently finished making sendmail 5.59 work here on altdorf, an HP 9000/350 running HP-UX 6.01. At this point the only problem I seem to have is with configuration freezing. When I run sendmail -bz, sendmail silently exits. This makes it seem like the freeze was successful, but when a later sendmail tries to read the frozen configuration it says "No local mailer defined" and dies with a segmentation error. Has anybody else encountered this behavior? It is obviously not critical; sendmail can function fine without a frozen .cf. But it would be nice to have. I have no idea what the problem might be. I don't know enough about the innards of a Bobcat.
allyn@hp-sdd.hp.com (Allyn Fratkin) (09/08/88)
In article <3077@hermes.ai.mit.edu>, wisner@altdorf.AI.MIT.EDU (Bill Wisner) writes: > When I run sendmail -bz, sendmail silently exits. This makes it seem > like the freeze was successful, but when a later sendmail tries to read > the frozen configuration it says "No local mailer defined" and dies > with a segmentation error. to fix the problem, remove the word "static" from the declaration of SymTab on or about line 38 of stab.c. this is due to a difference in the way arrays declared "static" are handled by the different (berkeley vs. system v) compilers. system v compilers put static data into the data section of the code file while static data goes into bss under berkeley. sendmail's freeze/thaw routines work by writing out and later restoring only the bss and dynamically allocated sections of the executing image. so the SymTab wasn't saved and it was empty in the new core image. -- From the virtual mind of Allyn Fratkin allyn%hpsdlx@sde.hp.com San Diego Division - or - Hewlett-Packard Company uunet!ucsd!hp-sdd!allyn