c to asm source lines ratio

[ca1]

i wonder how much it might be.
my guess is 4 for average short function about 20 lines x 80
characters.
i know it is very difficult to come up with some "average c code" and
"average asm code" .... nevertheless someone might have some
experience.
- thanks

[santosh]

ca1 wrote:

i wonder how much it might be.
my guess is 4 for average short function about 20 lines x 80
characters.
i know it is very difficult to come up with some "average c code" and
"average asm code" .... nevertheless someone might have some
experience.
- thanks

The answer would be different for different pieces of code and for CISC
vs. RISC architectures. Generally the assembler equivalent of a segment
of C code tends to be several times as long in terms of lines. Some
assembler dialects can have multiple instruction mnemonics in a single
line and of course C code can be written in wide ranging manner with
regard to vertical space. Compiler optimisation can dramatically affect
this ratio to IME.

[Walter Roberson]

In article <91**********************************@m36g2000hse. googlegroups.com>,
ca1 <xc*********@gmail.comwrote:

>i wonder how much it might be.
my guess is 4 for average short function about 20 lines x 80
characters.
i know it is very difficult to come up with some "average c code" and
"average asm code" .... nevertheless someone might have some
experience.

Completely unscientific experiment with a small number of small programs:
about 6 2/3.

--
"Do not on any account attempt to write on both sides of
the paper at once." -- Walter C. Sellar

[Walter Banks]

It is not a reliable metric. We have examples where several lines
of source yields one instruction and the converse is also true.

w..
ca1 wrote:

i wonder how much it might be.
my guess is 4 for average short function about 20 lines x 80
characters.
i know it is very difficult to come up with some "average c code" and
"average asm code" .... nevertheless someone might have some
experience.
- thanks