Compile Server in Debian

[Scura]

1 hour ago, nicolasg said:




2



I deleted the distro out of frustration xD so I can't provide more information at the moment, I made a post on pvpers, according to @Wizatek it was a problem with the version of gcc I was using (it was the version that came with Debian 11), try Centos 7 and gcc 4.8.5 and it kept giving the same problem, right now I'm trying with an iso that happened to me...




Still wrong boost version ... check if it is solved with the 1.67

[WhosUrDaddi]

1_65_0 should be the one you can use without changes to the code

[nicolasg]

it is very strange, in windows with boost 1.64 I have no problem compiling and making everything work. This is the result with centOS/gcc4.8.5/boost-1.65.0, I'll try with 1.67...

[nicolasg]

little update: I was able to compile perfectly using centOS 6.10/boost-1.65.1/gcc 4.4.7, thanks to /index.php?/profile/18894-desarija/&do=hovercard" data-mentionid="18894" href="/index.php?/profile/18894-desarija/" rel="">@Desarija, now my question is, is gcc really the problem?, because the command he use to compile boost is the same as I have I've been using this whole time...

[Scura]

8 hours ago, nicolasg said:




little update: I was able to compile perfectly using centOS 6.10/boost-1.65.1/gcc 4.4.7, thanks to /index.php?/profile/18894-desarija/&do=hovercard" data-mentionid="18894" href="/index.php?/profile/18894-desarija/" rel="">@Desarija, now my question is, is gcc really the problem?, because the command he use to compile boost is the same as I have I've been using this whole time...




Since you were using a wrong version of boost (which is the log in gdb too) probably the error may be related to this one ... i said it years ago in this conversation ... 

[Nikolee]

Im currently running into the same issue, turns out that the error is happening because of bad memory allocation from the memory pool..

I couldn't resolve it until now, but im sure it has something todo with boost and gcc.. if i remove the allocation from the pool it does work, but thats not the way..

[nicolasg]

16 hours ago, Nikolee said:




Im currently running into the same issue, turns out that the error is happening because of bad memory allocation from the memory pool..



I couldn't resolve it until now, but im sure it has something todo with boost and gcc.. if i remove the allocation from the pool it does work, but thats not the way..




he problem is gcc, it was difficult for me to install an older version, so I finally gave up. I use CentOS to compile. Then if I run everything in debian.

[Nikolee]

I was able to resolve it.. i can compile and run the server source on latest ubuntu or what-ever OS and also using latest boost version and C++17... 

But it was not that easy and i wouldn't recommend doing it if you dont have the experience or the knowledge.

+

[Rudi2404]

On 4/14/2022 at 10:16 PM, nicolasg said:




gdb currently return that, i'm completely lost at this point



2




I've now the same error on Helper, Connector and I think the other servers too.

 

Compiling machine: CentOS 7 x64

boost: 1.65.1

gcc: 4.8.5 20150623 (Red Hat 4.8.5-44) (GCC)

 

/index.php?/profile/5387-nikolee/&do=hovercard" data-mentionid="5387" href="/index.php?/profile/5387-nikolee/" rel="">@Nikolee: Can you post your solution, please?

 

Thanks to all for helping...

[Nikolee]

I can give you a hint but not my solution it was frustrating to do it and took me more than a week to resolve it.

 

If you update your compiler you should/need to update also the boost version and other libraries.

To migrate from boost-1-59 to 1-8X you need to do some changes on the networking stuff which is not really critical since mostly they changed the namespace and some constructors.

But changing this stuff won't make your server work, it should be able to compile but starting it will result in a segmentation fault.

The segmentation error happens because of bad memory allocation, you need to find out where it happens.

If you look at the trace exactly and you have knowledge then you should notice weird behavior, it has nothing to do with the memory pool from boost that works perfectly its just

code in the server which has different behavior in the newer compiler version.

I wouldn't recommend you to do this upgrade, because you cannot be sure about the behavior later on, we are talking here about the start of the server, but remember how many

lines of code the GS has, so if some crashes occur later on, you can't be sure if the crashes are because of the upgrade or some other stuff.

Note that if you do that stuff and make it work Unix you will have to do the same for windows but there are different problems because

its a different compiler..

For experienced developer this should be no issue and I hope that Gamigo is running their servers also on latest versions.. if not that's a really sad story..

 

One thing that is pretty cool after the upgrade is that I can use the latest C++ version now which makes it easier to write (good)code.

And also the performance increased heavily i could measure an boost of over 50% starting up the server, I have no data yet about the core performance but I think it should also be improved a lot.