| 1 |
-----BEGIN PGP SIGNED MESSAGE----- |
| 2 |
Hash: SHA1 |
| 3 |
|
| 4 |
Mike Frysinger wrote: |
| 5 |
> On Saturday 01 April 2006 17:17, Zac Medico wrote: |
| 6 |
>> Mike Frysinger wrote: |
| 7 |
>>> i thought about this some more ... why do we even care about the mtimes |
| 8 |
>>> during package merge ? if a package doesnt install any files into a |
| 9 |
>>> libdir, why should we bother running ldconfig ? if a user updated the |
| 10 |
>>> dirs, then they can either run ldconfig or env-update ... |
| 11 |
>>> |
| 12 |
>>> so imo, the logic should be: |
| 13 |
>>> if srcroot is None: |
| 14 |
>>> <mtime checks> |
| 15 |
>>> else: |
| 16 |
>>> <libdir existence checks> |
| 17 |
>> Yeah, I think that's correct. Also, even though we don't need to check the |
| 18 |
>> mtimes when srcroot is defined, we can still update the mtimes in the |
| 19 |
>> mtimedb to the current values in order to avoid an unnecessary ldconfig run |
| 20 |
>> on the next env_update call. |
| 21 |
> |
| 22 |
> feel like implementing this ? that way we dont stop on each others feet again |
| 23 |
> and i dont have to waste time when you rewrite my work ;) |
| 24 |
> -mike |
| 25 |
|
| 26 |
While investigating this I've found that srcroot will actually be emptied of all files in some cases. For example, this happens when PORTAGE_TMPDIR is on the same root filesystem that is being merged to, and the files are merged via os.rename. In these cases, srcroot becomes useless (passing in the CONTENTS would do the trick though). Because of this, the srcroot parameter to env_update should be removed. |
| 27 |
|
| 28 |
Zac |
| 29 |
|
| 30 |
-----BEGIN PGP SIGNATURE----- |
| 31 |
Version: GnuPG v1.4.2.2 (GNU/Linux) |
| 32 |
|
| 33 |
iD8DBQFEMxxD/ejvha5XGaMRAih0AJ9cIl6peHsEN2/MmYQeRpNOvAEQXwCfVMhz |
| 34 |
QgJko+J7xgTgfXzzgGIYNsE= |
| 35 |
=EdVn |
| 36 |
-----END PGP SIGNATURE----- |
| 37 |
-- |
| 38 |
gentoo-portage-dev@g.o mailing list |
Archives