Use "gfloat" and "gdouble", rather than "float" and "double", as the
authorguy <guy@f5534014-38df-0310-8fa8-9805f1628bb7>
Wed, 24 Apr 2002 21:53:05 +0000 (21:53 +0000)
committerguy <guy@f5534014-38df-0310-8fa8-9805f1628bb7>
Wed, 24 Apr 2002 21:53:05 +0000 (21:53 +0000)
commit377b23a07510fa764ab43976988d17af7f3504e2
tree28f11dfa7bce80675d7d156dc6386188c4f061a8
parent855585bbcf6512fa0b74cef3dbc21eaab1ba887f
Use "gfloat" and "gdouble", rather than "float" and "double", as the
return types of the tvbuff accessors for floating-point types, to more
closely match the tvbuff accessors for integral types.

Fix an error in the code for fetching doubles on VAXes, and get rid of
unused union members on VAXes.

git-svn-id: http://anonsvn.wireshark.org/wireshark/trunk@5245 f5534014-38df-0310-8fa8-9805f1628bb7
epan/tvbuff.c
epan/tvbuff.h