Summary: | sys-apps/baselayout-1.12.14-r1: /etc/init.d/netmount needs to be updated to reflect portmap -> rpcbind change | ||
---|---|---|---|
Product: | Gentoo Linux | Reporter: | Jeroen Roos <bugzilla-gentoo> |
Component: | [OLD] baselayout | Assignee: | Gentoo's Team for Core System packages <base-system> |
Status: | RESOLVED WONTFIX | ||
Severity: | major | CC: | G.Jaekel |
Priority: | Normal | ||
Version: | unspecified | ||
Hardware: | All | ||
OS: | Linux | ||
Whiteboard: | |||
Package list: | Runtime testing required: | --- | |
Attachments: | This patch fixes the problem on my computer |
Description
Jeroen Roos
2011-04-06 18:36:21 UTC
Created attachment 268753 [details, diff]
This patch fixes the problem on my computer
except the rpcbind init.d script has a "provide portmap" which means no other script should need to change I stepped into exactly the same issue and workaround. And I confirm the reason traced down by Jeroen. A comment in /etc/init.d/netmount says: # Only try to mount NFS filesystems if portmap was started. # This is to fix "hang" problems for new users who do not # add portmap to the default runlevel. And by replacing portmap by rpcbind this strikes: portmap isn't started and the following lines will disable the "mount -a" for nfs and nfs4. Ok, just bumped into this one myself - so is this patch the correct fix or not? It did seem to fix it for now, but will this break something else later?? As the patch just changes the term 'portmap' by 'rpcbind' it have to be correct in it's scope and may not break more than before. An enhanced version will try to determine what program have to be used, e.g.: for PORTMAPPER in rpcbind portmap; do PORTMAPPER=$(which $PORTMAPPER 2>/null) && break; done [ -x "$PORTMAPPER" ] || eerror "no portmapper found" and use $PORTMAPPER afterwards. openrc has different syntax (not that i see anything wrong with existing code) |