Follow @Openwall on Twitter for new release announcements and other news
[<prev] [next>] [<thread-prev] [day] [month] [year] [list]
Date: Wed, 13 Jul 2016 13:10:55 -0700
From: Jorge Almeida <>
Subject: Re: clockspeed

On Wed, Jul 13, 2016 at 12:42 PM, Rich Felker <> wrote:
> On Wed, Jul 13, 2016 at 12:04:36PM -0700, Jorge Almeida wrote:
>> Has anyone compiled clockspeed against musl, in a x86_64?
> I'm guessing you mean DJB's package,

but the name is generic enough
> that I can't be sure. Next time try to be more clear.

I had no idea there was some other package with the same name.

>> If so, would you share conf-cc and conf-ld?
> I just did as a test and it has some obvious issues like missing and
> wrong prototypes for standard functions and "extern int errno;" rather
> than #include <errno.h> (the latter is DJB's willfully wrong code to
> make a point that he likes being wrong), but it compiled "ok" after
> changing error.h. No idea if it works.
> I didn't modify conf-cc or conf-ld at all because the native gcc
> targets musl, but if you wanted to cross-compile it would probably
> work putting the name of your cross compiler there, or the musl-gcc
> wrapper even.

I'm not cross-compiling, but the distro is not musl-based. Yes, I put
the wrapper in conf-cc and conf-ld.

I've been using clockspeed for years (with < 5ms discrepancy relative
to the NTP server, on crappy hardware). But I had it compiled against
dietlibc long ago (in x86), and now it doesn't compile against dietlib
nor against musl. Just wondering if x86_64 brings about some extra


Powered by blists - more mailing lists

Confused about mailing lists and their use? Read about mailing lists on Wikipedia and check out these guidelines on proper formatting of your messages.