On Wed, 16 Nov 2011 22:33:56 +0100 Bernhard Reutner-Fischer <[email protected]> wrote:
> On Nov 16, 2011 12:50 PM, "William Pitcock" <[email protected]> > wrote: > > > > By using a function table, we can more cleanly support new crypt > > implementations, such as SHA256 ($5$) and SHA512 ($6$). > > Bloat-o-meter output? > thanks, Switching to this way of doing crypt() adds 132 bytes of overhead total. I do not personally think 132 bytes is enough to be concerned about. petrie:~/uClibc$ ~/dev-src/linux/scripts/bloat-o-meter ./lib-origin-master/libcrypt-0.9.33-git.so ./lib/libcrypt-0.9.33-git.so add/remove: 1/0 grow/shrink: 1/0 up/down: 132/0 (132) function old new delta crypt 92 192 +100 crypt_impl_tab - 32 +32 William _______________________________________________ uClibc mailing list [email protected] http://lists.busybox.net/mailman/listinfo/uclibc
