Hello, Looking back in the mailing list archive...
On Wed, Oct 12, 2022 at 8:00 PM Nicolas Chautru <nicolas.chau...@intel.com> wrote: > +/* Helper macro for logging */ > +#define rte_acc_log(level, fmt, ...) \ > + rte_log(RTE_LOG_ ## level, RTE_LOG_NOTICE, fmt "\n", \ > + ##__VA_ARGS__) I noticed that this macro logs messages with a 6 value as logtype, registered under lib.hash (RTE_LOGTYPE_HASH). lib/log/rte_log.h:#define RTE_LOG_NOTICE 6U /**< Normal but significant condition. */ lib/log/rte_log.h:#define RTE_LOGTYPE_HASH 6 /**< Log related to hash table. */ Please fix this. Here is some suggestion: Looking at other RTE_LOG_REGISTER in this driver, both acc and vrb drivers are registered under a same logtype (which is probably not entirely broken, but quite confusing): $ git grep RTE_LOG_REGISTER_DEFAULT drivers/baseband/acc/ drivers/baseband/acc/rte_acc100_pmd.c:RTE_LOG_REGISTER_DEFAULT(acc100_logtype, NOTICE); # mapped to pmd.bb.acc drivers/baseband/acc/rte_vrb_pmd.c:RTE_LOG_REGISTER_DEFAULT(vrb_logtype, NOTICE); # mapped to pmd.bb.acc It seems 3 logtypes are used in this code. So this driver can use sub types, like: RTE_LOG_REGISTER_SUFFIX(acc100_logtype, acc100, NOTICE); # mapped to pmd.bb.acc.acc100 RTE_LOG_REGISTER_SUFFIX(vrb_logtype, vrb, NOTICE); # mapped to pmd.bb.acc.vrb RTE_LOG_REGISTER_SUFFIX(acc_common_logtype, common, NOTICE); # mapped to pmd.bb.acc.common -- David Marchand
