Arguments:

buffer

The handle of the buffer you want to log the message in, or NULL to use the default buffer that you specified earlier with slog2_set_default_buffer().

code

A user-specified code that you want to be associated with the message. The slog2 system doesn't interpret this code in any way.

severity

The severity level of this log item; one of:

• SLOG2_SHUTDOWN
• SLOG2_CRITICAL
• SLOG2_ERROR
• SLOG2_WARNING
• SLOG2_NOTICE
• SLOG2_INFO
• SLOG2_DEBUG1
• SLOG2_DEBUG2

For more information, see the entry for slog2_register().

format

A printf-formatted character string used to define the log message.

Additional arguments

An argument vector that specifies the arguments for the format string. You must use the appropriate SLOG2_FA_*() helper macro for parameters, as described below. The last parameter in the vector must be SLOG2_FA_END.

Library:

libslog2

Use the -l slog2 option to qcc to link against this library.

Description:

The slog2fa() function a printf-style format string along with its arguments specified in the argument vector. This function uses special formatting macros to ensure that va_args() doesn't trigger any interrupts. The format string isn't decoded before writing; it must be decoded during the time of the read.

You can use the following macros to build the parameters to slog2fa:

• SLOG2_FA_UNSIGNED( x )
• SLOG2_FA_SIGNED( x )
• SLOG2_FA_CHAR( x )
• SLOG2_FA_FLOAT( x )
• SLOG2_FA_DOUBLE( x )
• SLOG2_FA_STRING( x )
• SLOG2_FA_END — you must end the list with this macro

For example:

int8_t some_number = -1;

slog2fa( ..., "string:%s, some_number:%d", SLOG2_FA_STRING( "Hello world" ),
                                           SLOG2_FA_SIGNED( some_number ),
                                           SLOG2_FA_END );

Returns:

0 on success, or -1 if an error occurred.

Classification:

QNX Neutrino

Caveats:

Don't use double or float arguments in an interrupt handler, unless your code is compiled to use software floating-point emulation.