Grim state of libpthread (needs serious cleanup)

Bernhard Reutner-Fischer rep.dot.nop at gmail.com
Wed Mar 25 13:38:27 UTC 2009


On Wed, Mar 25, 2009 at 12:19:34PM +0100, Carmelo AMOROSO wrote:
>-----BEGIN PGP SIGNED MESSAGE-----
>Hash: SHA1
>
>rhabarber1848 wrote:
>> Bernhard Reutner-Fischer wrote:
>> 
>>> I think that improving NPTL would be a good thing, i suppose that
>>> we ultimately want to abandon the other impls in favour of it, at
>>> least mid- or long-term.
>> 
>> Hi,
>> 
>> would it then still be possible to use uClibc with Linux kernel 2.4.x?
>I'm wondering if does it make sense n 2009 still using kernel 2.4...
>just a my opinion.

my stance on 2.4 vs. 2.6 is that if there are regressions in 2.6 compared to
2.4 then 2.6 should be fixed ;P
>
>> Especially in the embedded sector kernel 2.4 is often used because it is
>> smaller than 2.6.
>> 
>embedded worlds is now moving to NAND, and size will not longer a big
>deal.

I disagree, size does matter if it increases for no good reason.

>      And what about performance ? O(1) scheduler that came with kernel
>2.6, just to talk about one of the most important change in 2.6 series ?
>should not be the time to upgrade ?

Indeed. And i do agree that 2.6 is too big, but that's not exactly news.


More information about the uClibc mailing list