| 1 |
-----BEGIN PGP SIGNED MESSAGE----- |
| 2 |
Hash: SHA256 |
| 3 |
|
| 4 |
On 11/09/12 12:43 PM, Zac Medico wrote: |
| 5 |
> On 09/11/2012 09:36 AM, vivo75@×××××.com wrote: |
| 6 |
>> Dunno where to place this request, but if we go for something |
| 7 |
>> like EJOBS could we also make it phase specific? So compile, |
| 8 |
>> install and test could have a different number of jobs running. |
| 9 |
>> Possibly three different variables that override a predefined |
| 10 |
>> EJOBS. |
| 11 |
> |
| 12 |
> Per-phase sounds a little to fine-grained. Instead, I'd suggest to |
| 13 |
> add an ELOADAVG variable that's analogous to make's --load-average |
| 14 |
> option. That should be enough to compensate for any differences |
| 15 |
> between phases. |
| 16 |
|
| 17 |
I personally wonder about why this would be necessary from the |
| 18 |
perspective of the user; if the user's system at emerge time can |
| 19 |
handle X concurrent processes per emerge-job , i don't see why it |
| 20 |
would matter what phase these jobs would be launched from. |
| 21 |
|
| 22 |
At the ebuild level, certainly, but that's one of the reasons for |
| 23 |
EJOBS in the first place, so that it can be overridden consistently |
| 24 |
within a phase, if necessary for the ebuild (regardless of build |
| 25 |
system type), right? |
| 26 |
|
| 27 |
-----BEGIN PGP SIGNATURE----- |
| 28 |
Version: GnuPG v2.0.19 (GNU/Linux) |
| 29 |
|
| 30 |
iF4EAREIAAYFAlBPbLEACgkQ2ugaI38ACPA1qAD/bvjy7aB6nk5YboJHnCpQ8C56 |
| 31 |
QolKD9BPHL9eN8Xf41oA/iTZU+tyB+BDl+woZAlVGbaa6AR2r6Qp8rwOzkUWSwV/ |
| 32 |
=FhKc |
| 33 |
-----END PGP SIGNATURE----- |
Archives