Re: configure optimization feature update


Xu, Dongxiao <dongxiao.xu@...>
 

-----Original Message-----
From: Khem Raj [mailto:raj.khem@...]
Sent: Thursday, June 16, 2011 9:29 AM
To: Xu, Dongxiao
Cc: Richard Purdie (richard.purdie@...);
yocto@...
Subject: Re: [yocto] configure optimization feature update

On Wed, Jun 15, 2011 at 5:57 PM, Xu, Dongxiao <dongxiao.xu@...>
wrote:
Hi Richard,

Recently I was doing the "configure optimization" feature and collecting data
for it.

The main logic of this feature is straight forward:

1. Use the diff file as autoreconf cache. (I use command: "diff -ruN
SOURCE-ORIG SOURCE", here "SOURCE-ORIG" is the source directory before
running autoreconf, while "SOURCE" is the directory after running autoreconf).
2. Add SRC_URI checksum for all patches of the source code.
3. Tag each autoreconf cache file with ${PN} and the SRC_URI checksum of
source code and all patches.
4. If the currently SRC_URI checksum matches the cached checksum, then we
can patch the cache instead of running "autoreconf" stage.
The autoconf'ing is sort of arbitrary at the moment. Depending on what is
staged the results may vary. So some way of creating a cache is nice since it
can be used to verify if package rebuild happens with same configure variables
or not. Sometimes we have seen some packages are build differently first time
since many packages are not staged.


I did some testings for sato build, the result is not as good as we expected:

On a server build machine (Genuine Intel(R) CPU @ 2.40GHz, 2 sockets with 6
core each and hyperthreading, thus 24 logical CPUs in all, 66G memory):

w/o the optimization:
real    83m40.963s
user    496m58.550s
sys     329m1.590s

w/ the optimization:
real    79m1.062s
user    460m58.600s
sys     347m42.120s

It has about 5% performance gain.

I also tested the patch on a desktop core-i7 machine (Intel(R) Core(TM) i7
CPU 870 @ 2.93GHz, 4 core 8 logical CPU, 4G memory):

w/o the optimization:
real    105m25.436s
user    372m48.040s
sys     51m23.950s

w/ the optimization:
real    103m38.314s
user    332m35.770s
sys     49m4.520s

It only has about 2% performance gain.

The result is not encouraging.

There are also some other things we need to take into consideration for this
feature:

1. If add this feature, the first build time should be longer than current since it
needs to build the autoreconf cache.
2. Maintainers needs to maintain the SRC_URI checksums not only for source
code, but also all its patches. For some recipes, it has more than 20 patches,
which needs assignable maintenance effort.

Yeah thats definite pain.

3. How to distribute the caches will be a problem. The total size of such cache
is about 900M (before compression) and 200M (after compression). Since the
size is not small, distributing it with Poky source code doesn't make sense. On
another aspect, we can use something like "sstate". But since we already have
caches of sstate, I think it is not necessary for us to enable another similar
cache mechanism with little improvement.
hmm this is a real problem and probably the perf killer. I wonder why the sizes
are so big.
I diff the source files before and after autoreconf and using the generated patch as our autoreconf cache. For a certain source code, for example connman, such diff file will be as large as 3M, and for a sato build, we have about 300 recipes need autoreconf.

Thanks,
Dongxiao


Therefore my opinion is we may give up this feature. What's your comments
and suggestions?

Thanks,
Dongxiao

_______________________________________________
yocto mailing list
yocto@...
https://lists.yoctoproject.org/listinfo/yocto

Join yocto@lists.yoctoproject.org to automatically receive all group messages.