On Apr. 16, 2009, 21:23 +0300, "J. Bruce Fields" <bfields@xxxxxxxxxxxx> wrote: > On Thu, Apr 16, 2009 at 02:13:16PM -0400, Steve Dickson wrote: >> Benny Halevy wrote: >>>> Benny Halevy wrote: >>>>> Steve, please review the following patch that add support >>>>> for controlling the nfsv4 minor version support via >>>>> /proc/fs/nfsd/versions. >>>>> >>>>> [PATCH 1/4] utils/nfsd: fix -N optarg error printout >>>>> This patch fixes an existing bug. >>>> This is a bug... and has been committed... >>>> >>>>> [RFC 2/4] utils/nfsd: add support for minorvers4 >>>>> Under-the-cover support for minorvers4 >>>> I see you let minorvers4 default to zero, which means 4.1 >>>> support is off by default. Why? As long as we have away to >>>> turn of 4.1 processing (i.e. your 4/4 patch), then I see >>>> no reason we should have the support enabled by default. > > Was there a typo there? You ask why 4.1 support is off, then say you > see "no reason we should have the support enabled by default." Those > two statements seem to agree? Did you mean "no reason we should > not..."? (that is what I assumed.) > >>> I was also thinking about using the new nfs-utils with old kernels. >>> Though, these should not puke on seeing [-+]4.1, they'll >>> just interpret it as enabling/disabling v4. >>> I'll test that... >> I just did... Using the '-N 4.1' flag to rpc.nfsd on an older >> kernel simply turns off all v4 processing... which is fine, IMHO... > > If you want v4 running, but don't trust v4.1 yet, and want to use a > mixture of new and old kernels--what configuration will you use? It's a bit hackish, but I think that prepending 4.1 to 4 should do the trick. Writing "+2 +3 -4.1 +4" will allow you to disable 4.1 while enabling 4.0 on all kernels, old and new. "+2 +3 +4.1 +4" would enable 4.0, and 4.1 on kernels that support it. Benny > > --b. -- To unsubscribe from this list: send the line "unsubscribe linux-nfs" in the body of a message to majordomo@xxxxxxxxxxxxxxx More majordomo info at http://vger.kernel.org/majordomo-info.html
