Make bif and nif casing consistent (happi#29)

beam.asciidoc

@@ -213,7 +213,7 @@ again _{call,1,{f,4}}_.
 
 After the call x0 contains _id(B)_ and y0 contains _id(A)_,
 now we can do the addition: _{gc_bif,'+',{f,0},1,[{y,0},{x,0}],{x,0}}_.
-(We will go into the details of bif calls and gc later.)
+(We will go into the details of BIF calls and GC later.)
 
 === Dispatch: Directly Threaded Code
 
@@ -610,12 +610,12 @@ amount of work without using up its reductions.
 [WARNING]
 ====
 
-There are some bifs that can run for a long time only using 1
+There are some BIFs that can run for a long time only using 1
 reduction, like +term_to_binary+ and +binary_to_term+. Try to make
-sure that you only call these biffs with small terms or binaries, or
+sure that you only call these BIFs with small terms or binaries, or
 you might lock up the scheduler for a very long time.
 
-Also, if you write your own Nifs, make sure they can yield and that
+Also, if you write your own NIFs, make sure they can yield and that
 they bump the reduction counter by an amount proportional to their
 run time.
 

calls.asciidoc

@@ -78,7 +78,7 @@ Shouldn't Code Loading come before Hot Code Loading? Or are the two topics not r
 
 
 In the Erlang Runtime System the code loading is handled by the
-code server. The code server will call the lover level bifs in the
+code server. The code server will call the lover level BIFs in the
 +erlang+ module for the actual loading. But the code server also
 determines the purging policy.
 

genop.tab

@@ -437,7 +437,7 @@ BEAM_FORMAT_NUMBER=0
 122: bs_save2/2
 123: bs_restore2/2
 
-# New GC bifs introduced in R11B.
+# New GC BIFs introduced in R11B.
 
 ## @spec gc_bif1 Lbl Live Bif Arg Reg
 ## @doc Call the bif Bif with the argument Arg, and store the result in Reg.
@@ -454,7 +454,7 @@ BEAM_FORMAT_NUMBER=0
 ##      for the result (saving Live number of X registers).
 125: gc_bif2/6
 
-# Experimental new bit_level bifs introduced in R11B.
+# Experimental new bit_level BIFs introduced in R11B.
 # NOT used in R12B.
 126: -bs_final2/2
 127: -bs_bits_to_bytes2/2

memory.asciidoc

@@ -420,8 +420,8 @@ executing Erlang code.
 
 Note that no Erlang process running on the same scheduler as the
 allocator may start executing Erlang code before the block is freed.
-This means that you can not use a temporary allocation over a bif
-or nif trap (yield).
+This means that you can not use a temporary allocation over a BIF
+or NIF trap (yield).
 
 In a default R16 smp system there is N+1 temp_alloc allocators where N
 is the number of schedulers. The temp_alloc uses the "A fit" (+af+)
@@ -469,7 +469,7 @@ except for some short lived or temporary data used by ets tables-
 
 ==== The driver allocator: driver_alloc
 
-The driver allocator is used for ports, linked in drivers and nifs.
+The driver allocator is used for ports, linked in drivers and NIFs.
 
 ==== The short lived allocator: sl_alloc