Developer forums (C::B DEVELOPMENT STRICTLY!) > CodeCompletion redesign

CodeCompletion slow on some machine

<< < (3/5) > >>

MortenMacFly:

--- Quote from: seb_seb0 on April 11, 2012, 08:58:13 pm ---I attach a patch with :
    - the modifications discussed above
    - some additional debugging code in NativeParser.cpp and codecompletion.cpp (can be deactivated).

--- End quote ---
Please remove the additional debugging from bug fixes. It is hard to maintain/follow otherwise. And also remove large portions of commented code. If you want us to follow your steps you shouldn't provide patches where 90% of it is actually not related to fixing a bug. ???

ollydbg:
@morten
The debug code in this patch can help us to profile the CC code. :)

MortenMacFly:

--- Quote from: ollydbg on April 15, 2012, 03:37:40 pm ---@morten
The debug code in this patch can help us to profile the CC code. :)

--- End quote ---
I guess using the profiler plugin and a testing environment you can do much better ;-)

ollydbg:

--- Quote from: MortenMacFly on April 15, 2012, 03:38:39 pm ---
--- Quote from: ollydbg on April 15, 2012, 03:37:40 pm ---@morten
The debug code in this patch can help us to profile the CC code. :)

--- End quote ---
I guess using the profiler plugin and a testing environment you can do much better ;-)

--- End quote ---
If I remember correctly, I did that and we also have discussion in our forum.
I once use profile plugin(basically use gcc pg options) to analysis CC code, but that was a bit mass report under Windows, I can't find much useful information.

seb_seb0:
@OllyDbg:
regarding the encoding problem: I have the same problem on a my computer. I believe it has something to do with the wxString::Printf %s format specifier, and your suggestion to add _T() is probably correct. I will correct the code.

@Morten
for the patch, it is not so easy to separate, because all the changes are located in 2 files. My SVN client (TortoiseSVN) does not allow me to select which changes in 1 file to use for the patch (or I do not know how to do it). The alternative would be to edit the patch manually, but it is error-prone
That beeing said, I totally agree with you : it would be better to provide 2 patches : one for the additionnal profiling, and one for the code change.
I will remove also the commented code. And I will post another patch, with hopefully a real fix.

Regarding the use of a profiler : I will try to use gprof and see if I get something useful out of it. I will do that on the last nightly, with just compiler and codecompletion activated. From my experience, a profiler is helpful with a code that you know. In the case of codeblocks, it will take me some time to just analyze the output of the profiler. This is mainly why I have tried to use manual targeted profiling. If the bottleneck is obvious, you can find it very quickly.
CB is not compiled with the compiler option -pg, correct ? I will need to recompile it.

Regarding the problem
it is apparently related to a code change in CB, and not in the CodeCompletion plugin (since same CodeCompletion code triggers the problem on last nightly, but not on the last debugger branch). It is perhaps thread related, perhaps not.

And please be patient : I can debug this problem only at work, and I have since more than 1 year a very high pressure on my project (reason why other projects seem abandonned, like the xpmeditor). I cannot afford to spend too much time in a row on this problem. I hope everybody can understand that.

Seb

Navigation

[0] Message Index

[#] Next page

[*] Previous page