This is the mail archive of the libc-alpha@sources.redhat.com mailing list for the glibc 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: PATCH: BZ 464: IA64 math functions aren't weak


On Thu, Oct 28, 2004 at 11:54:43AM -0700, Ulrich Drepper wrote:
> -----BEGIN PGP SIGNED MESSAGE-----
> Hash: SHA1
> 
> H. J. Lu wrote:
> 
> > We just want to make ia64 behave the same as other targets.
> 
> And I've told you that this has rather been an unnecessary than a
> correct and necessary change.  So stop it.  If a function's
> implementation is not correct or good, fix it.

People want to override the math functions in glibc not because they
are bad. I don't believe any OS vendor will provide many optimized
libm.a for every different flavor of a given target. There are always
be cases where a user wants some highly optimized math functions for
his/her specific target and/or environment to get the last bit of
floating point performance. Currently, people can do it on many targets,
except for ia64.


H.J.


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