This is the mail archive of the gcc@gcc.gnu.org mailing list for the GCC project.


Index Nav: [Date Index] [Subject Index] [Author Index] [Thread Index]
Message Nav: [Date Prev] [Date Next] [Thread Prev] [Thread Next]
Other format: [Raw text]

Re: Faster compilation speed


David S. Miller <davem@redhat.com> writes:
>   From: Dale Johannesen <dalej@apple.com>
>   Date: Wed, 14 Aug 2002 10:17:46 -0700
>   
>   And I know this is blindingly obvious, but RC takes an extra field (word,
>   probably) in each node.  I suspect this is going to eat up a lot of
>   whatever gain there might be.
>   
>My implementation (I posted the hacked up infrastructure patch the
>other day) used space which is currently empty alongside a bitfield
>in the rtx.

Yet another speed/space trade-off - most architecures are going to take
significantly longer to inc/dec a bitfield than they will doing 
an int. 

RC is obviously less expensive in space when the "things" are larger,
an extra int is horribly expensive when there are only a few words
in the struct, but harmless when there are thousands.

Which reminds me that one of the advantages of the "obstack" scheme 
was it tended to act as a "slab allocator" with relatively few 
chunks with lots of little things inside each chunk.


-- 
Nick Ing-Simmons
http://www.ni-s.u-net.com/


Index Nav: [Date Index] [Subject Index] [Author Index] [Thread Index]
Message Nav: [Date Prev] [Date Next] [Thread Prev] [Thread Next]