On Tue, 7 Aug 2012 08:34:07 -0500 Geanta Neag Horia Ioan-B05471 <B05471@xxxxxxxxxxxxx> wrote: > On Tue, 7 Aug 2012 09:54:14 +0300, Geanta Neag Horia Ioan-B05471 <B05471@xxxxxxxxxxxxx> wrote: > > On Tue, 7 Aug 2012 04:47:51 +0300, Phillips Kim-R1AAHA <R1AAHA@xxxxxxxxxxxxx> wrote: > > >> Also: if the algorithm name is really all that changes, can it be > >> done more space-efficiently with some string manipulation in the > >> algorithm registration loop?: If registering an "authenc" > >> algorithm, also register an "authencesn" version. > > > > Agree. Will come up with smth more elegant in v2. > > > > Looking closer at algorithm registration, talitos_alg_alloc uses shallow copies > of the templates in driver_algs[], so maybe sticking with the initial approach > is safer ? not sure what you mean by 'shallow copies', but the t_alg->algt = *template assignment should be copying the entire struct crypto_alg. There are lots more redundancies in driver_algs, e.g., all .type == ALG_TYPE_AEADs have the same .cra_type and entry points. So, ideally, as support for more algorithms are added, factoring out the the common assignments into talitos_alg_alloc should keep the size of driver_algs from exploding and becoming even more unmanageable. Kim -- To unsubscribe from this list: send the line "unsubscribe linux-crypto" in the body of a message to majordomo@xxxxxxxxxxxxxxx More majordomo info at http://vger.kernel.org/majordomo-info.html
