svn commit: samba r19602 - in branches: SAMBA_3_0_23/source/lib SAMBA_3_0_24/source/lib

tridge at samba.org tridge at samba.org
Tue Nov 7 09:48:48 GMT 2006


Metze,

 > I assume this has bugs, as tm->tm_mon == 12 is valid and the array which
 > uses it as index has only 12 members...
 > 
 > tridge: had you gave me a reason, why you used > 12 in the samba4 commit?

I use > 12 because a value of 12 doesn't cause a segv. The loop is:

	for (i = 0; i < tm->tm_mon; ++i)
		res += ndays[is_leap(tm->tm_year)][i];

so i goes to max 11 with tm_mon == 12. That is fine with the array
given.

Really what this means is the '31' for December is not actually needed
in the code, which makes sense, as you never need to know how many
days in December in this calculation (as its implied in the 365
value).

anyway, either > 11 or > 12 is fine, both will stop the potential
segv.

Cheers, Tridge


More information about the samba-cvs mailing list