monotone-devel
[Top][All Lists]
Advanced

[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

[Monotone-devel] Re: illegal instruction again


From: Robert Bihlmeyer
Subject: [Monotone-devel] Re: illegal instruction again
Date: Sun, 11 Apr 2004 13:25:57 +0200
User-agent: Gnus/5.1006 (Gnus v5.10.6) XEmacs/21.4 (Security Through Obscurity, linux)

Peter Simons <address@hidden> writes:

> You cannot say
>
>     CXXFLAGS=-O0 ./configure
>
> but have to use:
>
>     export CXXFLAGS=-O0
>     ./configure

Both these constructs will put CXXFLAGS into the environment of the
configure process, they are equivalent in this regard. (What configure
does with it in respect to its subprocesses is its business, but it
certainly can't tell the difference between the two cases above.)

~$ foo=bar sleep 100 &                  
[1] 3643
~$ tr \\0 \\n < /proc/3643/environ | grep '^foo' 
foo=bar
~$ export foo=baz
~$ sleep 100 &
[1] 3659
~$ tr \\0 \\n < /proc/3659/environ | grep '^foo'
foo=baz

> Oh, and about the '-msse2' flag: I wouldn't enable MSSE
> support, just because some header file exists! Personally, I
> would never enable it automatically but provide an
> --enable-msse2 option to configure. If one wants to check, I
> think
>
>     grep msse /proc/cpuinfo
>
> is a better test -- at least on Linux. But all that fails
> miserable when you are cross-compiling.

Seconded. My preferences are:
1. run-time detection                                   [tricky]
2. --with-msse2=(yes|no|auto) where auto guesses from
   the environment, and no is the default               [easy]
3. don't optimise                                       [trivial]
4. guess from the environment by default                [easy]


Footnotes:
[1] if we ignore the effect on commands started from this shell after configure

-- 
Robbe

Attachment: pgpCvGyFGKS83.pgp
Description: PGP signature


reply via email to

[Prev in Thread] Current Thread [Next in Thread]