Re: Real simple cache that removes most of the lookups in mcstrans

[James Antill <jantill@redhat.com>]

[-- Attachment #1: Type: text/plain, Size: 1006 bytes --]

On Wed, 2006-05-17 at 12:19 -0400, Daniel J Walsh wrote:

> Only reason strdup fails is ENOMEM.  With ENOMEM you are almost 
> garanteed you are going to crash anyways.  gstrdup does a exit when it 
> runs out of memory.
> So we can messy up the code with a lot of checks that end up doing 
> little.  Your choice.

 Well, it doesn't currently fail that way ... and it's relying on a side
effect that sometimes happens.
 Also the grep's I can see show the malloc/realloc/strdup failures being
handled, so without bugs in the application layer ENOMEM doesn't
currently mean a crash.

 I'd be happy to do the diffs to add an internal
xstrdup()/xmalloc()/etc. which calls abort(), if that's the route we
want to go. As you say, this would significantly reduce the failure
paths ... although it might get libselinux banned from some
applications.
 I'll also volunteer to do an audit and add the failure paths, if we
want to handle ENOMEM.

-- 
James Antill
<james.antill@redhat.com>


[-- Attachment #2: This is a digitally signed message part --]
[-- Type: application/pgp-signature, Size: 191 bytes --]