Most popular GMP downloads
Andreas Enge
andreas.enge at inria.fr
Thu Apr 24 08:38:17 UTC 2014
On Thu, Apr 24, 2014 at 10:00:06AM +0200, Richard Biener wrote:
> I checked the configure script and it accepts the .0 releases
> (but complains "buggy but acceptable"). Similar for GMP where
> it accepts 4.2.3 but complains until you have at least 4.3.2.
Okay.
> There is also no easy way (working across gmp/mpfr/mpc releases)
> to check for the version of an in-tree variant without first
> building it (thus we apply no version check there).
Mpc contains the following in src/mpc.h (since mpc-0.6 in 2009):
/* Define MPC version number */
#define MPC_VERSION_MAJOR 1
#define MPC_VERSION_MINOR 0
#define MPC_VERSION_PATCHLEVEL 2
#define MPC_VERSION_STRING "1.0.2"
> I'll see to update the versions used for GCC trunk.
Excellent! You may wish to have a look at the bug report to coordinate your
efforts. A paragraph I just posted which is relevant to the discussion
(although as I am speaking about mpc, maybe on the wrong list here):
"Even if you are not using always the latest version of the multiprecision libraries, the ftp server and download script should nevertheless be updated from time to time. Moreover, at least for mpc (I cannot speak for the others, but I think there it also holds) you should always use the latest bug fix release in a release series. Since the 1.0.0 version, we maintain a separate branch for the 1.0 series; so 1.0.2 is strictly a bug fix release, and new functionality will go into the 1.1 release series. After the 1.1 release, we will continue to maintain the 1.0 branch for a while. Details have not been decided yet, but I think that in any case, once the 1.2 release is out, the 1.0 branch will not be maintained any longer. At this point in time, you should definitely switch to a newer release in the download script.
Right now, the 0.8 and 0.9 versions are unmaintained; we started a more conservative policy of releasing and backporting bug fixes in 2012 with 1.0 (as indicated by the version number...)."
This goes against your wishes of updating rather infrequently...
Andreas
More information about the gmp-devel
mailing list