trevor> Thank you for reading it. Yes, I had originally changed those too,
trevor> but decided to leave them alone after getting "doesn't need config"
trevor> messages.
[Just an information note, no action needed]
That's because 2.0.XX 'mkdep.c' has a bug in it. It recognizes CONFIG_*
variables only when they appear in lines that start with #if.
It doesn't recognize CONFIG_* in constructs like this:
/* include/linux/if_frad.h */
/* devices which this FRAD is slaved to */
struct device *master[CONFIG_DLCI_MAX];
short dlci[CONFIG_DLCI_MAX];
For the 2.0.XX tree I think we should let sleeping bugs lie. If some
code truly fails to pick up a definition for CONFIG_DLCI_MAX here then
the C compiler will tell us anyways.
For 2.1.XX, 'make dep' no longer issues these diagnostics at all --
'make checkconfig' does this job now, and it handles this case correctly.
Michael Chastain
<mailto:mec@shout.net>
"love without fear"