[PATCH] [RFC] cryptodev: replace LIST_END enumerators with APIs

[Kusztal, ArkadiuszX]

> -----Original Message-----
> From: Morten Brørup <mb at smartsharesystems.com>
> Sent: Thursday, September 5, 2024 5:09 PM
> To: Akhil Goyal <gakhil at marvell.com>; dev at dpdk.org
> Cc: thomas at monjalon.net; Marchand, David <david.marchand at redhat.com>;
> hemant.agrawal at nxp.com; anoobj at marvell.com; De Lara Guarch, Pablo
> <pablo.de.lara.guarch at intel.com>; Trahe, Fiona <fiona.trahe at intel.com>;
> Doherty, Declan <declan.doherty at intel.com>; matan at nvidia.com;
> g.singh at nxp.com; fanzhang.oss at gmail.com; jianjay.zhou at huawei.com;
> asomalap at amd.com; ruifeng.wang at arm.com;
> konstantin.v.ananyev at yandex.ru; Nicolau, Radu <radu.nicolau at intel.com>;
> ajit.khaparde at broadcom.com; rnagadheeraj at marvell.com; mdr at ashroe.eu
> Subject: RE: [PATCH] [RFC] cryptodev: replace LIST_END enumerators with APIs
> 
> > +++ b/app/test/test_cryptodev_asym.c
> > @@ -581,7 +581,7 @@ static inline void print_asym_capa(
> >  			rte_cryptodev_asym_get_xform_string(capa-
> >xform_type));
> >  	printf("operation supported -");
> >
> > -	for (i = 0; i < RTE_CRYPTO_ASYM_OP_LIST_END; i++) {
> > +	for (i = 0; i < rte_crypto_asym_op_list_end(); i++) {
> 
> > +++ b/lib/cryptodev/rte_crypto_asym.h
> > +static inline int
> > +rte_crypto_asym_xform_type_list_end(void)
> > +{
> > +	return RTE_CRYPTO_ASYM_XFORM_SM2 + 1; }
> > +
> >  /**
> >   * Asymmetric crypto operation type variants
> > + * Note: Update rte_crypto_asym_op_list_end for every new type added.
> >   */
> >  enum rte_crypto_asym_op_type {
> >  	RTE_CRYPTO_ASYM_OP_ENCRYPT,
> > @@ -135,9 +141,14 @@ enum rte_crypto_asym_op_type {
> >  	/**< Signature Generation operation */
> >  	RTE_CRYPTO_ASYM_OP_VERIFY,
> >  	/**< Signature Verification operation */
> > -	RTE_CRYPTO_ASYM_OP_LIST_END
> >  };
> >
> > +static inline int
> > +rte_crypto_asym_op_list_end(void)
> > +{
> > +	return RTE_CRYPTO_ASYM_OP_VERIFY + 1; }
> 
> I like the concept of replacing an "last enum value" with a "last enum function"
> for API/ABI compatibility purposes.
> 
> Here's an idea...
> 
> We can introduce a generic design pattern where we keep the _LIST_END enum
> value at the end, somehow marking it private (and not part of the API/ABI), and
> move the _list_end() function inside the C file, so it uses the _LIST_END enum
> value that the library was built with. E.g. like this:
> 

Why asym crypto API does need these ENDs at all? Will any PMD or the user ever or use it? Sym crypto does not have this at all, as well as rte_crypto_asym_ke_type in asym crypto.


> 
> In the header file:
> 
> enum rte_crypto_asym_op_type {
> 	RTE_CRYPTO_ASYM_OP_VERIFY,
> 	/**< Signature Verification operation */ #if RTE_BUILDING_INTERNAL
> 	__RTE_CRYPTO_ASYM_OP_LIST_END /* internal */ #endif }
> 
> int rte_crypto_asym_op_list_end(void);
> 
> 
> And in the associated library code file, when including rte_crypto_asym.h:
> 
> #define RTE_BUILDING_INTERNAL
> #include <cryptodev/rte_crypto_asym.h>
> 
> int
> rte_crypto_asym_op_list_end(void)
> {
> 	return __RTE_CRYPTO_ASYM_OP_LIST_END;
> }