On 01/25/2018 09:42 PM, Stephen Hemminger wrote:
On Thu, 25 Jan 2018 17:00:43 +0000
Andrew Rybchenko <arybche...@solarflare.com> wrote:
diff --git a/drivers/net/sfc/efsys.h b/drivers/net/sfc/efsys.h
index c7a54c3..8dd225e 100644
--- a/drivers/net/sfc/efsys.h
+++ b/drivers/net/sfc/efsys.h
@@ -26,6 +26,7 @@
#include <rte_io.h>
#include "sfc_debug.h"
+#include "sfc_log.h"
#ifdef __cplusplus
extern "C" {
@@ -721,7 +722,7 @@ typedef uint64_t efsys_stat_t;
#define EFSYS_ERR(_esip, _code, _dword0, _dword1) \
do { \
(void)(_esip); \
- RTE_LOG(ERR, PMD, "FATAL ERROR #%u (0x%08x%08x)\n", \
+ SFC_GENERIC_LOG(ERR, "FATAL ERROR #%u (0x%08x%08x)", \
(_code), (_dword0), (_dword1)); \
_NOTE(CONSTANTCONDITION); \
} while (B_FALSE)
Off topic, but why is this header file having C++ wrapper?
It is driver private, and driver is always built with C.
In this particular case it is just few lines which, as I understand,
never hurt. So, it is better to have and not think about it.
I have no strong opinion on it.
