[ECOS] compiling options

[ECOS] compiling options

[Rafael Rodríguez Velilla]

[-- Warning: decoded text below may be mangled, UTF-8 assumed --]
[-- Attachment #1: Type: text/plain, Size: 548 bytes --]

  I have seen some options that are default in eCos while compiling for
which I can't find documentation. these are:
-fdata-sections
-fno-rtti
-fno-exceptions
-fvtable-gc
-finit-priority
-Wl,--gc-sections

  Can somebody tell me what are they for or where is the documentation
about those?

  I have also read that ffunction-sections is used but I don't
understand why. It seems to generate worse code, which is its advantage?

--
Rafael Rodríguez Velilla        rrv@tid.es
Telefónica I+D          http://www.tid.es
Telf: +34 - 91 337 4270

RE: [ECOS] compiling options

[Dan Conti]

[-- Warning: decoded text below may be mangled, UTF-8 assumed --]
[-- Attachment #1: Type: text/plain, Size: 923 bytes --]

>   I have seen some options that are default in eCos while compiling for
> which I can't find documentation. these are:
> -fdata-sections
> -fno-rtti
> -fno-exceptions
> -fvtable-gc
> -finit-priority
> -Wl,--gc-sections
>
>   Can somebody tell me what are they for or where is the documentation
> about those?

http://www.gnu.org/manual/

check out the manual for gcc, it has information about various command line
options.

>
>   I have also read that ffunction-sections is used but I don't
> understand why. It seems to generate worse code, which is its advantage?

-ffunction-sections causes functions to be put in their own section. used in
conjunction with --gc-sections, you get selective linking, so that functions
that aren't called aren't linked in to your final image.

-Dan

>
> --
> Rafael Rodríguez Velilla        rrv@tid.es
> Telefónica I+D          http://www.tid.es
> Telf: +34 - 91 337 4270
>
>
>
>

Re: [ECOS] compiling options

[Rafael Rodríguez Velilla]

Thank you all for the information. I thought  that the man pages that where
installed while I compiled the toolchain should be the most uptodate of my
particular tools.

Re: [ECOS] compiling options

[Hugo 'NOx' Tyson]

danc@iobjects.com (Dan Conti) writes:
> >   I have also read that ffunction-sections is used but I don't
> > understand why. It seems to generate worse code, which is its advantage?
> 
> -ffunction-sections causes functions to be put in their own section. used in
> conjunction with --gc-sections, you get selective linking, so that functions
> that aren't called aren't linked in to your final image.

That's right.

It will also mean that individual .o files, and libraries &c *are* larger.
But that has nothing to do with the *code* in the final image.

Is that why it looked like worse code?

	- Huge

Re: [ECOS] compiling options

[Jonathan Larmour]

[-- Warning: decoded text below may be mangled, UTF-8 assumed --]
[-- Attachment #1: Type: text/plain, Size: 955 bytes --]

Rafael Rodríguez Velilla wrote:
> 
>   I have seen some options that are default in eCos while compiling for
> which I can't find documentation. these are:
> -fdata-sections
> -fno-rtti
> -fno-exceptions

See http://gcc.gnu.org/onlinedocs/gcc_3.html for these and
-ffunction-sections

The -fno-exceptions is the opposite of -fexceptions - it disables C++
exceptions support.

> -fvtable-gc

Garbage collect unused C++ vtable entries.

> -finit-priority

C++ constructor priorities. Required for eCos.

> -Wl,--gc-sections

"-Wl," means it is passed to the linker. See:
http://sources.redhat.com/binutils/docs-2.10/ld_2.html#SEC2

>   I have also read that ffunction-sections is used but I don't
> understand why. It seems to generate worse code, which is its advantage?

Why do you think so?

Jifl
-- 
Red Hat, Rustat House, Clifton Road, Cambridge, UK. Tel: +44 (1223) 271062
Maybe this world is another planet's Hell -Aldous Huxley || Opinions==mine