Re: /usr/include/linux and stuff

["Stephen J. Turnbull" <nospam [at] sk.tsukuba.ac.jp>]

>>>>> "Al" == Al Hudson <eah106 [at] york.ac.uk> writes:

    Al> On Wed, 16 Feb 2000, Stephen J. Turnbull wrote:

    Al> Kernel includes shouldn't be included in user-land
    Al> compilation. Simple as that?!

Oh.  Then it's not glibc's fault.  Case closed?

    Al> Once glibc is compiled, the headers should match the
    Al> library. It also makes it a complete arse if you i) retarget
    Al> compilers, ii) have multiple glibc/libc, iii) have multiple
    Al> kernel source trees.

Agreed.  So what?  You can't solve this problem without (a) freezing
the kernel headers permanently (Linus won't find that acceptable ;) or
(b) tightly coupling glibc with the kernel the way (eg) modules and
(better eg) pcmcia-cs's cardmgr is (at least under Debian).  Users
won't find that acceptable, I suspect.

    >> If your program uses both glibc and direct calls to the kernel,
    >> you probably want to compile with the same set of headers as
    >> your glibc, and hope the kernel can take care of itself.

    Al> Erm, yuck? ;)))

Tell David Hinds that.  Does the implementor of PCMCIA really have a
choice?  He can't use glibc-isms, his code has to run on BSD systems
too.

-- 
University of Tsukuba                Tennodai 1-1-1 Tsukuba 305-8573 JAPAN
Institute of Policy and Planning Sciences       Tel/fax: +81 (298) 53-5091
_________________  _________________  _________________  _________________
What are those straight lines for?  "XEmacs rules."
---------------------------------------------------------------------
Sheffield Linux User's Group - http://www.sheflug.co.uk
To unsubscribe from this list send mail to
- <sheflug-request [at] vuw.ac.nz> - with the word 
 "unsubscribe" in the body of the message. 

  GNU the choice of a complete generation.