[Mono-dev] mcs patch to say goodbye to SeekableStreamReader

[Atsushi Eno]

Hi,

I just commited and reverted the patch - gmcs seems pretty complex
on those tokenization matter than mcs. I'll try to make those
changes on gmcs directly.

Atsushi Eno

Miguel de Icaza wrote:
> Hello,
> 
>> I agree that xtoken() could be performance sensitive, but on the
>> other hand, current code moves back the stream, which results in
>> extraneous parse on every parenthesized expressions. Actually
>> as mentioned later, the patched code seems faster. Note that
>> SeekableStreamReader has to call encoding GetCharCount() when
>> we reposition the stream.
> 
> Good point.  Thanks for the benchmarks, lets get this patch into SVN.
> 
>> Actually the resulting performance is better than existing code:
>>
>> existing code at *best*:
>> real    0m2.711s
>> user    0m0.080s
>> sys     0m0.050s
>>
>> Total memory allocated: 34835 KB
>>
>> patched code at *worst*:
>> real    0m2.635s
>> user    0m0.020s
>> sys     0m0.030s
>> Total memory allocated: 34083 KB
>>
>> Most of the memory consumption difference comes from
>> SeekableStreamReader:
> 
> Lovely, thanks for the patch and the measurements.
> 
>> ########################
>>      731 KB Mono.CSharp.SeekableStreamReader::.ctor(StreamReader)
>>          731 KB      363 System.Char[]
>>   Callers (with count) that contribute at least for 1%:
>>          363  100 % Mono.CSharp.SeekableStreamReader::.ctor(Stream,Encoding)
>>
>> Atsushi Eno