On 10/20/2017 01:32 PM, Florian Weimer wrote:
> ----- Original Message -----
>> + #if defined(_SC_IOV_MAX) /* IRIX, MacOS X, FreeBSD, Solaris, ... */
>> +- __svc_maxiov = sysconf(_SC_IOV_MAX);
>> ++ {
>> ++ /*
>> ++ * some glibc (e.g. 2.26 in Fedora 27 beta) always
>> ++ * return -1
>> ++ */
>> ++ int i = sysconf(_SC_IOV_MAX);
>> ++ if (i != -1 && i > __svc_maxiov)
>> ++ __svc_maxiov = i;
>> ++ }
>> + #endif
>> + return true;
>> + }
>
> Uh-oh, is this really necessary?
You said it'd be a while before fixed packages make it into Fedora 27.
Without this we're dead in the water.
That aside, I don't understand what the concern is. This is merely
defensive and largely or effectively a no-op.
>
> We have almost fixed the glibc bug.
>
> Anyway, what's the difference between libntirpc and libtirpc? Is there an
> expectation that libntirpc eventually replaces libtirpc?
libtirpc is frozen. (lib)ntirpc is a _new_ library with many new
features and bug fixes needed by nfs-ganesha. No, I don't believe anyone
has any expectation that it will ever replace libtirpc.
(This was all discussed at length back when the libntirpc was originally
packaged in Fedora.)
>
> Do you regularly import fixes from the glibc code into libntirpc?
>
I don't know. That's a question for the ntirpc developers (cc'd)
--
Kaleb
------------------------------------------------------------------------------
Check out the vibrant tech community on one of the world's most
engaging tech sites, Slashdot.org! http://sdm.link/slashdot
_______________________________________________
Nfs-ganesha-devel mailing list
Nfs-ganesha-devel@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/nfs-ganesha-devel
