On Wed, 2 May 2018 10:40:08 +0100 Ferruh Yigit <ferruh.yi...@intel.com> wrote:
> On 5/1/2018 7:56 PM, Thomas Monjalon wrote: > > 01/05/2018 19:28, Ferruh Yigit: > >> --- a/lib/librte_ethdev/rte_ethdev.c > >> +++ b/lib/librte_ethdev/rte_ethdev.c > >> -static int ethdev_logtype; > >> - > >> -#define ethdev_log(level, fmt, ...) \ > >> - rte_log(RTE_LOG_ ## level, ethdev_logtype, fmt "\n", ## __VA_ARGS__) > >> +int ethdev_logtype; > >> > > [...] > >> --- a/lib/librte_ethdev/rte_ethdev.h > >> +++ b/lib/librte_ethdev/rte_ethdev.h > >> +extern int ethdev_logtype; > >> + > >> +#define RTE_ETHDEV_LOG(level, fmt, ...) \ > >> + rte_log(RTE_LOG_ ## level, ethdev_logtype, fmt "\n", ## __VA_ARGS__) > > > > Why moving this macro in the header file? For using it in inline functions? > > Probably worth an explanation in the commit message. > > Yes mainly for RTE_PMD_DEBUG_TRACE usage in header file. The ones in > RTE_ETH_VALID_PORTID_... macros and inline functions, which I didn't update in > this set. > > I will update the commit log. It would be great to eliminate all usage of PMD log level (and maybe even EAL) in a future release.
