man portage

<snip>
              package.provided
                     A list of packages (one per  line)  that  portage
should
                     assume  have  been  provided.  Useful for porting to
non-
                     Linux systems.  Portage will  not  attempt  to  update
a
                     package  that  is  listed  here  unless  another
package
                     explicitly requires a version that is newer than what
has
                     been  listed.   Basically,  it's a list that replaces
the
                     emerge --inject syntax.

                     For example, if you manage your own copy of a 2.6kernel,
                     then  you  can tell portage that
'sys-kernel/development-
                     sources-2.6.7' is already taken care of and it should
get
                     off your back about it.

                     Virtual  packages  (virtual/*) should not be specified
in
                     package.provided.  Depending on the type of  virtual,
it
                     may  be  necessary  to  add an entry to the virtuals
file
                     and/or add a package that satisfies a  virtual  to
pack-
                     age.provided.

</snip>

Also, if you've already emerged it (maybe even manually how you make/make
install), you might try emerge --noreplace <package>, which should add the
ebuild to the world file (but not re-emerge the package itself).  If the
package is already in the world file, then this option will do nothing.

On Dec 7, 2007 9:50 PM, Michael Sullivan <[EMAIL PROTECTED]> wrote:

> A lot of the bigger packages (qt, mythtv, etc. etc) tend to lock up my
> PCs' hard drives while they're emerging.  I use FEATURES=keepwork on all
> my boxes, and I can usually go
> into
> /var/tmp/portage/<whatever-class>/<whatever-package/work/<whatever-package>
> and issue a make and then make install after I reboot the machines.  How can
> I tell portage that the package is installed in this manner?  That the
> package is indeed installed?
>
> --
> [EMAIL PROTECTED] mailing list
>
>


-- 
- Mark Shields

Reply via email to