| 1 |
On N, 1970-01-01 at 00:00 +0000, Tobias Klausmann wrote: |
| 2 |
> Hi! |
| 3 |
> |
| 4 |
> On Wed, 29 Aug 2012, William Hubbs wrote: |
| 5 |
> > On Wed, Aug 29, 2012 at 01:57:48PM +0200, Tobias Klausmann wrote: |
| 6 |
> > > As a first crude datapoint, I compared the build times |
| 7 |
> > > (configure+make) of udev-171-r6 and -188 on our dev Alpha. This |
| 8 |
> > > is a machine that's on the speedier side of off-mainstream |
| 9 |
> > > architecures, but as a datapoint, it should be enough. |
| 10 |
> > |
| 11 |
> > No, not exactly, because udev-188 doesn't build systemd. We use |
| 12 |
> > specific targets in the Makefile to avoid doing that. |
| 13 |
> |
| 14 |
> Amending my earlier post then, here are the numbers for the |
| 15 |
> 171 build and the target-specific build of 188: |
| 16 |
> |
| 17 |
> ver (1) (2) (1+2) |
| 18 |
> udev-171-r6: 15s / 71s / 86s; 1m26s |
| 19 |
> udev-188 : 28s / 78s / 116s; 1m56s |
| 20 |
> |
| 21 |
> Much less difference. Still, I figure _really_ slow |
| 22 |
> machines/arches might be in more pain. I'll run my test with a |
| 23 |
> Geode-based x86 later this week. |
| 24 |
|
| 25 |
Geode LX700 (433MHz) with 256MB RAM |
| 26 |
MAKEOPTS=-j2 (single core system) |
| 27 |
gcc (Gentoo 4.5.2 p1.1, pie-0.4.5) 4.5.2 |
| 28 |
|
| 29 |
ebuild prepare done before as well. |
| 30 |
|
| 31 |
1. time ebuild foo configure — real time value |
| 32 |
2. time ebuild foo compile — real time value |
| 33 |
|
| 34 |
ver (1) (2) (1+2) |
| 35 |
udev-171-r6: 47.2s / 4m36.4 / 5m23.6 |
| 36 |
udev-188 : 69.6s / 6m27.2 / 7m36.8 |
| 37 |
|
| 38 |
Personally of course don't care the slightest of this time. |
| 39 |
|
| 40 |
|
| 41 |
Mart |
Archives