From mboxrd@z Thu Jan 1 00:00:00 1970 From: Andrew Cagney To: jtc@redback.com Cc: gdb@sourceware.cygnus.com Subject: Re: savestring() Date: Wed, 21 Mar 2001 15:59:00 -0000 Message-id: <3A9FC4DC.334CBD54@cygnus.com> References: <5m1ysgopq8.fsf@jtc.redback.com> X-SW-Source: 2001-03/msg00007.html "J.T. Conklin" wrote: > > Lint complaints about signed/unsigned conversions in the typical use > of savestring: ARI just complains about it :-) > x->foo = savestring(bar, strlen(bar)); > > Is there any reason why we shouldn't change the type of savestring's > (likewise msavestring) second argument from int to size_t? > > Also, since the above is the same as: > > x->foo = xstrdup(bar); > > Is there any reason why we shouldn't change it? It's a bit odd that > half the code uses savestring() and the other half uses xstrdup(). >From the TODO file. -- Replace savestring() with something from libiberty. An xstrldup()? but that would have different semantics. -- There may be more on this in the GDB and OpenBSD e-mail archives. I sent an e-mail to the OpenBSD group asking if there should be a strldup() like function - the strl*() functions originated from someone in OpenBSD land. Anyway, yes, some instances of savestring should be xstrdup() but others are not so should probably get ``X'' rated treatment :-) As for size_t. I have no idea, does libiberty manage to use size_t in its headers? If it does then, why not :-) Andrew