tracing: have trace_printk process arguments before storing
trace_printk records its arguments in the internal ring buffer for them to be processed later. The mapping between the format string for the particular instance of a trace_printk call and the arguments is maintained using the instruction pointer internally in the kernel. However, this makes it harder for an external tool to process this data, which is necessary when the same data is logged to the STM and needs to be processed externally. Change-Id: Ifa2ae5fe1f1fd62c85fac432a581737450781f88 Signed-off-by: Pushkar Joshi <pushkarj@codeaurora.org> (cherry picked from commit c86e25f0f775d7effaf192ded6fb7a6885694d7b)
This commit is contained in:
committed by
Rohit Vaswani
parent
2d168136d4
commit
715363e77e
@@ -486,7 +486,7 @@ do { \
|
||||
__attribute__((section("__trace_printk_fmt"))) = \
|
||||
__builtin_constant_p(fmt) ? fmt : NULL; \
|
||||
\
|
||||
__trace_bprintk(_THIS_IP_, trace_printk_fmt, ##args); \
|
||||
__trace_printk(_THIS_IP_, trace_printk_fmt, ##args); \
|
||||
} else \
|
||||
__trace_printk(_THIS_IP_, fmt, ##args); \
|
||||
} while (0)
|
||||
|
||||
Reference in New Issue
Block a user