| 1 |
On Wed, Jun 8, 2011 at 3:40 AM, "Paweł Hajdan, Jr." <phajdan.jr@g.o> |
| 2 |
wrote: |
| 3 |
> On 6/7/11 9:53 PM, Matt Turner wrote: |
| 4 |
>> Building 400~ locales is not fun on mips when building stages. |
| 5 |
> |
| 6 |
> Do you have some data to quantify "not fun"? How long does it take? |
| 7 |
|
| 8 |
To build glibc |
| 9 |
|
| 10 |
- once for the n32 ABI with only the en_US.UTF-8 locale: 1h 15m |
| 11 |
- three times for o32/n32/n64 ABIs with only en_US.UTF-8: 4h 50m |
| 12 |
- three times for o32/n32/n64 ABIs with all locales: 6h |
| 13 |
|
| 14 |
I can reconfirm how long it takes just by running locale-gen on all locales, |
| 15 |
but according to my numbers it takes 1h 10m to generate all the locales. |
| 16 |
|
| 17 |
The o32/n32/n64 configuration is what I've been planning to ship as a |
| 18 |
compromise between shipping only a single ABI and a set of stages for every |
| 19 |
ABI, but each ABI adds significant time, and n32 replaces o32 so I'll just |
| 20 |
plan to drop o32. |
| 21 |
|
| 22 |
So roughly the time required to build all the locales is equal to the time |
| 23 |
to build glibc for a single ABI. |
| 24 |
|
| 25 |
>> No user has a need for more than some small subset of the total |
| 26 |
>> available locales. |
| 27 |
>> |
| 28 |
>> I filed bug [1] to request the ability to select locales in catalyst |
| 29 |
>> spec files, but no responses after six months -- which is totally |
| 30 |
>> typical of catalyst bugs. |
| 31 |
>> |
| 32 |
>> I commented in bug [2] suggesting that we perhaps reduce the default |
| 33 |
>> locale.gen to only 'en_US.UTF-8 UTF-8' or some other limited subset. |
| 34 |
> |
| 35 |
> It makes sense to me, especially for "exotic" arches. It seems that |
| 36 |
> locale-gen is cheap anyway on x86 and amd64, where it probably matters |
| 37 |
most. |
| 38 |
|
| 39 |
Good, thanks for your feedback. |
| 40 |
|
| 41 |
Matt |
Archives