[Patch] Step2: don't use _M_out_lim in filebuf, have filebuf non unified, etc...

[Paolo Carlini]

Nathan Myers wrote:

>On Tue, Jun 24, 2003 at 12:17:26AM +0200, Paolo Carlini wrote:
>
>>Of course many further refinements are possible but the below
>>brings us to a fairly stable, clean and fast streambuf.
>>
>Once we get the streambuf inlines shuffled around between
>std/std_streambuf.h and bits/streambuf.tcc, we'll be ready to run
>some races.  Eat our dust, stdio!  :-)
>
Eh, eh...

Anyway, *right now*, on my P4, for 9876
(see http://gcc.gnu.org/ml/libstdc++/2003-04/msg00245.html
for background), five consecutive runs of iostreams:

1.560u 0.280s 0:01.85 99.4%     0+0k 0+0io 216pf+0w
1.480u 0.380s 0:01.94 95.8%     0+0k 0+0io 216pf+0w
1.440u 0.400s 0:05.56 33.0%     0+0k 0+0io 216pf+0w
1.590u 0.270s 0:02.25 82.6%     0+0k 0+0io 216pf+0w
1.500u 0.370s 0:03.96 47.2%     0+0k 0+0io 216pf+0w

five of stdio_unlocked:

1.910u 0.430s 0:02.39 97.9%     0+0k 0+0io 215pf+0w
2.010u 0.360s 0:02.37 100.0%    0+0k 0+0io 215pf+0w
1.890u 0.420s 0:06.44 35.8%     0+0k 0+0io 215pf+0w
1.990u 0.370s 0:02.37 99.5%     0+0k 0+0io 215pf+0w
2.070u 0.320s 0:04.17 57.3%     0+0k 0+0io 215pf+0w

not bad, not bad...
Paolo.